Skip to main content

Full text of "The Origin of the Indo-Iranians"

See other formats


The Origin of the 
Inao-Iranians 




Elena E. Kuz'mina 

Edited byj.P. Mailory 



LEIDEN INDO-EUROPEAN ETYMOLOGICAL DICTIONARY SERIES 



The Origin of the IndoTranians 



Leiden Indo-European 
Etymological Dictionary Series 



Edited by 
Alexander Lubotsky 



VOLUME 3 



The Origin of the 
Indo-Iranians 



By 

Elena E. Kuz'mina 

Edited by 

J. P. Mallory 



6* ■* ' 



• S 

''681' 



BRILL 



LEIDEN • BOSTON 

2007 



Translated by S. Pitina and P. Prudovsky. 

With support of the Russian Institute for Cultural Research, Moscow, and Harvard University. 

This book is printed on acid-free paper. 



ISSN: 1574-3586 
ISBN: 978 90 04 16054 5 

Copyright 2007 by Koninklijke Brill NV, Leiden, The Netherlands. 
Koninklijke Brill NV incorporates the imprints Brill, Hotei Publishing, 
IDC Publishers, Martinus Nijhoff Publishers and VSP. 

All rights reserved. No part of this publication may be reproduced, translated, stored in 
a retrieval system, or transmitted in any form or by any means, electronic, mechanical, 
photocopying, recording or otherwise, without prior written permission from the publisher. 

Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV 
provided that the appropriate fees are paid directly to The Copyright Clearance Center, 
222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. 
Fees are subject to change. 

PRINTED IN THE NETHERLANDS 



Dedicated to the blessed memory of my teachers Mikhail and Igor D 'yakonov. 

Their methods of research, broad-minded historical approach, 

strict objective evaluation of facts, constant adherence to principles, 

and their benevolence were a guiding star for me. 



CONTENTS 

Introduction xi 

PART ONE 
THE ANDRONOVO CULTURAL ENTITY 

Chapter 1: History of Research on the Andronovo Culture 3 

Chapter 2: Methodological Aspects of Ethnocultural Reconstruction 9 

Chapter 3: Classification of Sites and the Primary Features of 

Andronovo Unity 17 

Petrovka 19 

Alakul' 20 

Fedorovo 23 

Sol'-Iletsk 26 

Kozhumberdy 27 

Atasu 27 

Amangel'dy 28 

Tautary 29 

Semirech'e 29 

Chapter 4: Settlements and Domestic Architecture 31 

Settlements and houses 31 

House-plan 36 

House Type I 40 

House Type II 60 

House Type III 66 

Chapter 5: Ceramics 67 

Petrovka 68 

AlakuF 69 

Fedorovo 71 

Mixed types 74 

Alekseevka 75 

Dongal 77 

Dandybay 78 

Chapter 6: Mining, Metallurgy and the Metal Industry 85 

Chapter 7: Textiles and Dress 101 

Chapter 8: Transport 107 

Vehicles 107 

Chariots of the Eurasian steppe 109 

Cheek-pieces 115 

Frontlet 131 

The tactics of chariot warfare and horse riding 131 



CONTENTS Vlll 

Chapter 9: Economy 141 

Farming 141 

Stockbreeding 146 

The Indo- Iranian economy 157 

Chapter 10: The Ethnogenesis of the Indo-Iranians 

and the Ethnic Attribution of the Andronovo Culture 163 

The retrospective approach 163 

The evidence of material culture 164 

Chapter 11: Verification of the Hypothesis 169 

Anthropological evidence 169 

Toponymic evidence for the Indo-Iranian homeland 173 

Indo- Iranian traditions on the homeland 174 

Art and mythology 175 

Some mythological representations 183 

Chapter 12: Mortuary Practice 185 

Chapter 13: Indo-Iranian Contacts with Other Linguistic Groups 199 

Indo-Iranians and Finno-Ugrians 199 

Indo-Iranians and Greeks 204 

Chapter 14: Conclusions to Part One 205 

PART TWO 

THE MIGRATIONS OF TRIBES AND THEIR CULTURES 

IN CENTRAL ASIA 

Chapter 15: Cultures of Central Asia in the 4th-3rd Millennia BC 211 

Chapter 16: The Indo- Aryan Migration and the First Stage of the 

Andronovo Migration to the South 217 

The problem of migration 217 

Migration in the Eurasian Steppe and neighboring regions 220 

The first stage of Andronovo migration 220 

Chapter 17: The Agricultural Tribes of South Central Asia 

in the 2nd Millennium BC 225 

Chapter 18: The Occupation of Central Asia by Pastoral Tribes 229 

The first Indo- Aryan migration to the south 229 

The second stage of Andronovo migration 234 

Chapter 19: The Settlement of Pastoral Tribes in Central Asia 237 

The Tazabagyab culture: The Aral Sea Littoral variant 238 

The Tazabagyab culture: The Lower Zeravshan variant 240 

Sites of the Central Zeravshan 240 

The Tashkent Oasis 241 

The Timber-grave culture 241 

Fedorovo type sites of the Andronovo culture 242 

Andronovo sites of the Fedorovo type in Kirgizia 242 

The Semirech'e-Fergana type 243 

Sites of the North-eastern Tian-Shan 246 

Sites of southern Fergana 247 

Fergana valley: Late sites of the Kayrak-Kum type 248 

Chapter 20: Relation of the Andronovans with the Population of 

Xinjiang and Central Asia 251 



IX CONTENTS 

Chapter 21: Cultures of Northern Bactriain the Late Bronze Age 267 

BMAC sites in Uzbekistan 267 

Northern Bactria and Tadzhikistan 271 

BMAC sites in Tadzhikistan 271 

Bishkent-Vaksh culture 275 

The Andronovo culture in Tadzhikistan 278 

Andronovo sites of the Pamirs 285 

The Andronovo culture in southern Bactria 286 

The culture of the Barbarian Occupation Period 289 

Chapter 22: Trans-Caspia and Turkmenia 291 

PART THREE 

THE GENESIS OF THE DIFFERENT BRANCHES 

OF INDO-IRANIANS 

Chapter 23: Modern State of the Problem of Indo-Iranian Origins 297 

Chapter 24: The Genesis of the Dards and Nuristani 307 

Chapter 25: The Genesis of the Indo- Aryans 321 

Is Vedic archaeology possible? 322 

The Bactria-Margiana Archaeological Culture 323 

The Andronovo culture 325 

Settlements and dwellings 326 

Ceramics 327 

Metal articles 328 

Horses and chariots in Hindustan 329 

Burial rite 339 

The Andronovo burial rite 340 

Burial rites in Baluchistan 342 

Burial rites in the Harappan civilization 342 

Burials in the post-Harappan cultures of North- Western India 343 

Anthropological and genetic data 345 

PART FOUR 
THE GENESIS OF THE IRANIANS 

Chapter 26: The Genesis of the Iranians 349 

The history of research on the Timber-grave culture 349 

The culture of the Eurasian steppes in the Final Bronze Age 358 

The genesis of the West Iranians 367 

The genesis of Eastern- Iranian nomadic peoples: 

the Scythians and the Saka 379 

The anthropological data 383 

The formation of the Scythian complex 388 

The genesis of the Eastern Iranians of South Central Asia, 

Afghanistan and Baluchistan 413 

The formation of the farming culture of the Iranian-speaking peoples 

in North Central Asia 436 

The Chust culture 437 

The Burgulyuk culture 441 

The problem of the date of Zarathustra 448 



CONTENTS X 

Conclusion 451 

Post Scriptum 458 

APPENDICES 

Appendix I: The Chronology of the Andronovo Culture 459 

The methods of determining chronology 459 

The new radiocarbon dates of the steppe cultures 460 

The European line of synchronization 461 

The South-Asian line of synchronization 462 

The Chinese line of synchronization 464 

Conclusion 465 

Appendix II: Radiocarbon dates 467 

BIBLIOGRAPHY 477 

MAPS 583 

FIGURES 609 

INDEX 731 



INTRODUCTION 

In the middle of the 2nd millennium BC Egyptian civilization reached its zenith 
under the pharaohs of the 1 8th dynasty, the Hittite kingdom arose in Asia Minor, 
the Achaeans created the Mycenaean civilization in Greece, the Monteoru cul- 
ture flourished along the Danube, central Europe was occupied by tribes of the 
Unetice culture, the early farmers of the Anau culture were erecting their citadels 
in western Central Asia, the rulers of the Shang dynasty established their power 
in China, and in India the Harappan civilization was in decline. But what was 
happening at this time in the vast expanses of the Eurasian steppes that spread 
unbroken from the Danube to the Great Wall of China? Here, in the middle of 
the 2nd millennium BC, two gigantic cultural spheres were formed with no par- 
allels in the Old World: in the west the Timber-grave (Srubna) culture stretched 
from the Danube to the Ural, and in the east the Andronovo culture extended 
from the Ural to the Yenisei (Map 1). And in the Near East the Indo-Iranians, a 
branch of the Indo-European language family, are mentioned for the first time. 

The time when Indo-Iranians are first recorded in cuneiform texts, the 
16th(?)-15th centuries BC, provides a terminus ante quern for the separation of 
the Indo-Iranians from the other Indo-European peoples. Proper names of the 
dynastic rulers of the Hurrian kingdom of the Mitanni in northern Mesopotamia 
and Syria are mentioned in written documents of the 17th-16th centuries. In the 
14th century BC Prince Sativasa with a Mitanni detachment escaped from Arrap- 
cha to the Hittite king Supilluliuma and concluded a treaty. This was guaranteed 
by an oath to the Hittite and Hurrian gods where four ancient Indo-Iranian 
gods — Indra, Varuria, and the Nasatya twins — are also mentioned. The Mitanni 
Kikkuli wrote a treatise on training stallions to be harnessed to chariots, in which 
the terms connected with horse-breeding and training were Indo-Iranian. 

This treatise was translated from the Hurrian language in the Hittite kingdom 
and Assyria. This has prompted some scholars to assign a special role to the 
Indo-Aryans in the dispersal of horse-breeding and chariots in the Near East 
(D'yakonov 1956; Kammenhuber 1961; Mayrhofer 1966, 1974; Salonen 1955- 
1956; Thieme 1960; Ivanov 1968). According to A. Kammenhuber and M. 
Mayrhofer the Arya arrived in Mitanni apparently somewhat before the middle 
of the 2nd millennium BC (rather than the mid-16th century BC). The question 
of what branch of Indo-Iranians the Near Eastern Arya belonged to has been a 
matter of debate. The point of view expressed by P. Thieme, M. Mayrhofer, and 
V. V. Ivanov prevails: the Arya of Mitanni were already separated Indo-Aryans 
or bearers of a special Aryan dialect, although their movement probably involved 
related Proto- Iranians. I. M. D'yakonov (1995: 129) suggested that they could be 
a third group of Indo-Iranians, the Dards and Nuristani. This evidence shows 
that, in the first place, before the middle of the 2nd millennium BC, Indo- 
Iranians had not only separated from other Indo-Europeans but also formed a 
community of separate dialects; in the second place, a special role in Indo- 



Xll INTRODUCTION 

Iranian culture belonged to horse-breeding and chariot tactics in battle, i.e., the 
period of the last contacts between Indo-Iranians in their homeland is set before 
the 16th- 15th centuries BC by the spread of chariots in the Old World and, 
consequently, the area of Indo-Iranian dialect contacts should fall within the 
region where we find developed horse-breeding and chariots. 

Two centuries have passed since W. Jones (1799) and A. Schlegel pronoun- 
ced on the relationship between the ancient Indians, Greeks and Romans and 
proposed an Indo-European language family. The original homeland of the Indo- 
Iranians was localized variously in India, Central Asia, the Pamirs, the Arctic, on 
the Danube, the north Pontic steppes, Iran and the Near East. The published 
criticism of these various hypotheses in the works by E. A. Grantovsky (1970), 
B. G. Gafurov (1972), J. P. Mallory (1973; 1998a; Mallory and Mair 2000), and 
E. Bryant (1999) makes a detailed review of these different points of view 
unnecessary. We may simply note that both in linguistics and archaeology there 
are two major competing hypotheses under serious discussion concerning the 
origins of the Indo-Iranians. 

T. V. Gamkrelidze and V. V. Ivanov (1972: 19-25; 1979: 33-37; 1980a: 64- 
71; 1980b; 1981a: 11-33; 1981b: 80-92; 1984) have proposed a new reconstruc- 
tion of the Indo-European proto- language which emphasized the kinship between 
Proto-Indo-European and the Kartvelian and Semitic languages, a reconstructed 
economic and cultural vocabulary more advanced than previously proposed, 
names of southern flora and fauna, and a homeland in a mountainous landscape. 
This and other linguistic evidence provided the basis for their transferring the 
Indo-European homeland from Europe to the Near East and they localized it in 
the 4th millennium BC before its disintegration in northern Asia Minor, whence 
Hittite-Luvians and Greeks made a short move to the west, Proto-Indo-Iranians 
advanced slightly to the east, to the northern part of the Iranian plateau, and from 
there some Indo-Aryans moved to the west into Mitanni, while others migrated 
east to India. They also argued that the Tocharians and the bearers of the ancient 
European dialects (Italics, Celts, Germans, Baits, Slavs and later also early 
Iranians) moved in subsequent waves during the 3rd and 2nd millennia BC 
through Central Asia into the northern Caspian and further west to the Volga and 
north Pontic. At the end of the 3rd millennium BC they spread across Europe 
from the area of the Pit-grave (Yamnaya) culture and again appeared in the Near 
East (the migration of the so-called "Sea Peoples"). The last to take the same 
route from the territory of Iran through Central Asia were the Eastern Iranians 
(Scythians and Sarmatians). 

V. V. Ivanov and T. V. Gamkrelidze' s proposals are now a focus of dis- 
cussion (D'yakonov 1982; Merpert 1985; Kuz'mina 1980b; 1994; Lelekov 1982; 
Grantovsky 1998; Mallory 1998a). J. Nichols (1996) locates the Indo-European 
homeland in Central Asia (Bactria), from whence she proposes a migration to the 
north Pontic took place. But the evidence of archaeology excludes such a 
movement of Indo-Europeans (Mallory 1998). 

The other major theory presumes an Indo-Iranian homeland in Europe. 
European linguists at the beginning of the 20th century proposed that the Indo- 
Iranians had separated at the end of the 3 rd / beginning of the 2 nd millennium BC, 
and their partial withdrawal from the homeland was through the Eurasian steppes 
in Central Asia, and further to the south, to India and Iran. However, a number of 



INTRODUCTION Xlll 

authors admitted that some Iranians, ancestors of the Scythians and probably the 
Cimmerians, remained in Europe. Others have suggested that the Scythians 
returned to Europe from Kazakhstan in the 1st millennium BC. Opinions differ 
also about the route of Indo-Iranian migration from a south Russian homeland or 
(at least partially) through the Caucasus (P. Kretschmer, E. A. Grantovsky, K. 
Jettmar, W. Brandenstein, R. Ghirshman, M. N. Pogrebova, P. Bosh-Gimpera) or 
through Central Asia. This point of view was held by W. Geiger (1882), O. 
Schrader (1901), E. Meier, V. Bartold, E. Benveniste (1966), V. Pisani, G. Ca- 
meron, A. Meillet (1938), A. Christensen (1943), E. Herzfeld (1947), and such 
specialists in Indo-European and Indo-Iranian languages as V. Georgiev (1958), 
G. Morgenstierne (1973), W. Porzig (1964), Th. Burrow (1973; 1976), W. Bran- 
denstein (1948; 1962), H. Bailey (1955; 1957), R. Hauschild (1962), R Frye 
(1972), M. Mayrhofer (1966; 1974), M. Boyce (1975), J. Harmatta (1981), and 
A. Parpola (1973; 1988). Various views were expressed by the Russian scholars 
V. V. Struve (1955), I. M. D'yakonov (1956; 1958; 1995; 1996), I. Aliev (1960), 
M. M. D'yakonov (1961), I. M. Oransky (1963; 1979a; 1979b), V. I. Abaev 
(1965; 1972; 1981), G. M. Bongard-Levin (Bongard-Levin and Iljin 1969; Bon- 
gard-Levin and Grantovsky 1983), E. A. Grantovsky (1970; 1998), T. Ya. Eliza- 
renkova (1972; 1989; 1995; 1999), B. G. Gafurov, (1972), M. A. Dandamaev 
(Dandamaev and Lukonin 1980), and until recently V. V. Ivanov (Ivanov and 
Toporov 1960: 10-22; Ivanov 1963: 11-18). V. Georgiev (1958: 280-282) 
proposed that the earliest Indo-Iranians could be associated with the Pit-grave 
culture. M. M. D'yakonov (1961) placed the Indo-Iranian homeland in south-east 
Europe to the east of the Tripolye culture. E. A. Grantovsky (1960: 351-357) 
held that "the epoch of the Pit-grave culture can correspond to the common 
Aryan period." V. I. Abaev (1965: 134-135) thought the Timber- grave culture to 
be Iranian. M. M. D'yakonov (1961: 40-42, 64) and E. A. Grantovsky (1970: 
359-360) directly connected Iranians or Indo-Iranians with the Timber-grave and 
Andronovo cultures but I. M. D'yakonov (1956: 124; 1960), W. Brandenstein 
(1962: 3), Th. Burrow (1973: 126), and M. Boyce (1975) connected them only 
with the Andronovo culture. I. M. D'yakonov (1970: 126; 1995; 1996), 
moreover, proposed that the Indo-Iranians left no later than in the second quarter 
of the 2 nd millennium BC. 

The conclusions of these linguists are accepted by the majority of archaeo- 
logists who specialize in the cultures of the Eurasian steppes and Central Asia of 
the 2nd millennium BC. They are S. P. Tolstov (1948: 68; 1962: 59), A. N. 
Bernshtam (1957: 18-19), S. S. Chernikov (1957; 1960: 112), K. F. Smirnov 
(1957b: 8-14;1964; Smirnov and Kuz'mina 1977), V. M. Masson (1959: 116- 
117), M. A. Itina (Tolstov and Itina 1960; Itina, 1979: 232-236), N. Ya. Merpert 
(1961: 172, 173; 1966: 149-160; 1974: 14), B. A. Litvinsky (1962: 291-295 
1963: 127-133; 1964; 1967: 122-126; 1981: 160-162), E. E. Kuz'mina (1963b: 
155-158; 1971a, 1971; 1972a; 1972b; 1974a, 1974b; 1981a; 1987a; 1988c; 1994 
1995; 1999; Smirnov and Kuz'mina 1977), K. V. Sal'nikov (1965: 347), A. M. 
Mandel'shtam (1966a: 258; 1968, ch.5, 6), K. A. Akishev (1973), V. F. Gening 
(1977b; 1985), M. N. Pogrebova (1977a: 133-140, 170), N. L. Chlenova (1980: 
66-67; 1983a; 1984; 1986), S. S. Berezanskaya (1982: 206-209), G. B. Zdano- 
vich (1992; 1995; 1999), and I. B. Vasil'ev et al. (1995). They hold that the 
archaeological evidence of the Bronze Age cultures of the Eurasian steppes does 



XIV INTRODUCTION 

not contradict this linguistic hypothesis and it is supported by the archaeological- 
ly established genetic connections between the Iranian-speaking Scythians, 
Sarmatians and Saka and the earlier Timber-grave and Andronovo cultures. 

Other cultures have been mentioned within the context of Indo-Iranian orig- 
ins and expansions, e.g., the Abashevo culture (Pryakhin 1977: 134-7) and the 
Catacomb culture (Klein 1980; 1984: 32-5). N. F. Guseva (1977: 27-51; 1987: 
46) assigned the Indo-Iranians to the Timber-grave and Abashevo cultures, and 
later the Andronovo, and stressed their connection with the Slavs. A. M. Man- 
del'shtam (1968: 131-141) conducted a systematic analysis of the funeral prac- 
tice of the Bishkent (Vakhsh) culture and demonstrated specific corresponden- 
ces with Indo- Aryan practices. He viewed the Bishkent culture as cattle raising, 
coming from the north-west in transit to India, and he noted its similarity to the 
Andronovo culture. B. A. Litvinsky (1964: 158; 1967: 122-126) connected this 
culture with the Nuristani languages and showed its analogies in Swat. E. E. 
Kuz'mina (1972 a: 134-143; 1972b: 116-121; 1974: 188-193; 1975: 64-7) em- 
phasized the Indo-Iranian attribution of the culture, its connection with Swat and 
Gomal and the participation of the Zamanbaba and Andronovo components in its 
formation. C. Silvi Antonini (1986) spoke against this, justly mentioning the ab- 
sence of catacombs in Swat, but new discoveries in Tadzhikistan now confirm 
the correctness of this hypothesis (Vinogradova and Kuz'mina 1986: 199). This 
does not, by any means, exclude connections between Swat and the agricultural 
cultures of Bactria (Sarianidi 1987) and Iran (Stacul 1970; Pogrebova 1977a: 
145). 

Some specialists in Iranian archaeology have suggested that the Indo-Iranian 
homeland was on Iranian territory. Some suggest that the makers of grey ware 
ceramics in the last quarter of the 2nd millennium BC were Indo-Iranian 
speakers. This oft-discussed ware is supposed to have originated in western Iran 
and Central Asia, and there is a continuity of ceramic tradition from the 3rd 
millennium BC (Hissar-III, Shah-Tepe). But the origin of the groups of Iranians, 
the Saka, Sarmatians and Scythians in the steppes, the Bactrians, Sogdians, etc. 
in Central Asia, is in general indisputable. This hypothesis is shared by L. 
Vanden Berghe (1964: 37), C. Young (1965: 72; 1967), R. Dyson (1965; 1967), 
J. Deshayes (1969), K. Jettmar (1956: 327-342), who initially considered Andro- 
novans Indo-Iranians but later regarded them to be Finno-Ugrians following V. 
N. Chernetsov. 

The association of Iranians with grey ware has been repeatedly criticized by 
A. M. Mandel'shtam (1964: 192-194), E. E. Kuz'mina (1975c: 650), V. G. 
Lukonin (1977: 13-15, 18), M. N. Pogrebova (1977a: 8,16), G. Cleusiou (1982), 
I. N. Medvedskaya (1977: 169-175; 1978: 7-9, 14-18), and especially by E. A. 
Grantovsky (1981: 245-272; 1998: 37-123). These critics have shown that, first, 
grey ware was not continuous through all the periods of its existence; second, 
grey wares are not identical in their various local variants, and they are met in 
different archaeological contexts; and, finally, grey wares dominate in those 
areas of Iran, which, according to written sources, were populated by non- 
Iranians: Hurrians, Kassites, Gutians, Lullubi, and Elamites. Thus, it remains 
highly disputable whether the phenomenon of grey ware is directly connected 
with Indo-Iranian populations. This hypothesis incidentally contradicts the Gam- 
krelidze-Ivanov model as it proposes that Iranians migrated from east to west. 



INTRODUCTION XV 

R. Ghirshman (1977) placed the Indo- Aryan and Iranian homeland in south- 
east Europe, but proposed that the unity collapsed in the 4th millennium BC 
when the Indo- Aryans migrated to Mesopotamia and Iran. He dated the arrival of 
the Iranians from their homeland only at the turn of the 2nd- 1st millennia BC, 
connecting the Iranians with different cultural complexes (Sialk-A and B, Yaz 1) 
and sites of ancient Dahistan. (For a critical analysis of R. Ghirshman' s state- 
ments see Lelekov 1978: 220-226; for objections to claiming Iran and south Cen- 
tral Asia as the center of the appearance of the chariot see Kuz'mina (1980a: 
28)). I. N. Khlopin (1970a: 94-112; 1970b: 57-58; 1970c: 8-89; 1983: 125) cate- 
gorically insisted on the agricultural character of the Indo-Iranian economy, and 
admitted the existence of the Indo-Iranians in Iran and in south Central Asia 
already from the 4th millennium BC. He connected the genesis of the Andronovo 
culture with the Anau culture, wherein the Andronovans were outcasts forced to 
move to the north where they had to adopt a primarily cattle-breeding economy. 
In this way they introduced agriculture, metallurgical skills and the Iranian lan- 
guage to native hunting populations (Khlopin 1970c: 95, 98-99). This approach 
was opposed by Andronovo scholars (Kuz'mina 1972a: 137-142). V. I. Sarianidi 
(1977: 113, 143-150, 158; 1990: 95-102) connects the Iranians not with the grey 
ware culture(s) but with the ceramic complex of Bactria. He holds that Iranian- 
speaking tribes migrated to the east and north from Iran in the 2nd millennium 
BC (where the supposedly eastern- Khorasan culture has not yet been discovered) 
and he strongly rejects the ascription of the Andronovans to the Indo-Iranians 
(1977: 149). He also suggests that bearers of the Tillya-Tepe painted pottery 
experienced a secondary migration; the question of the homeland of the Saka- 
Scythians is not addressed. V. I. Sarianidi' s statements were opposed by A. A. 
Askarov (1977: 156), who suggests, after V. M. Masson, that the creation of the 
agricultural culture of Bactria was the result of migration from south 
Turkmenistan. In his more recent work V. I. Sarianidi (1999), following T. V. 
Gamkrelidze and V. V. Ivanov, has traced some connections between the 
Bactria-Margiana Archaeological Complex (BMAC) and both the Aegean and 
Anatolia, localizing the Indo-Iranian homeland in the latter, and proposing 
Bactria as the homeland of Zoroastrianism. This hypothesis has been severely 
criticized (Mallory 1998b). The recent work of S. A. Grigor'ev, another 
proponent of the Gamkrelidze-Ivanov hypothesis, titled (in Russian) The Earliest 
Indo-Europeans. An attempt at historical reconstruction, Chelyabinsk, 1999, is 
beyond scientific criticism as it has serious methodological mistakes in the 
analysis of the archaeological material. 

Thus, the hypotheses that both the grey ware and the Bactrian ceramics of the 
2nd millennium BC are connected with the Iranians is neither proved nor 
generally accepted. Comparison of the wheel-made pottery of Achaemenid Iran 
with ceramics of south Central Asia shows a different genesis of ceramic 
traditions (Kuz'mina 1969; 1971c: 171-178; 1972a: 143-146; Sarianidi 1977: 
116, 151; Cattenat and Gardin 1977: 225-246), and it may be concluded that 
wheel-made pottery (as opposed to hand-made pottery) cannot be considered as a 
true ethnic indicator. 

Thus, no single hypothesis from the competing linguistic and archaeological 
solutions to the location of the Indo-Iranian homeland has been proved at pres- 
ent. Consequently, the task before us is a thorough comparison of the linguistic 



XVI INTRODUCTION 

and archaeological data and an assessment of the ethnic attribution of the 
Andronovo culture. But prior to this we need to analyze the material itself: while 
the Andronovo culture was identified more than 80 years ago it has never been 
the subject of a monograph. A second aim of this book then is partially to fill this 
gap by compiling a complete database drawn from the whole Andronovo area 
and to reconstruct its material culture, economy and cultural type. 

The Eurasian steppes of the 2 nd millennium BC witnessed the process of the 
final establishment of a pastoral economy. The third most important task of 
research is to explain the specific features of the steppe economy and the 
dynamics of its development. Without resolving these issues one cannot 
understand the reasons for the transition of the steppe population to mass 
nomadism at the beginning of the I s ' millennium BC, the processes of ethnic 
migrations during the 2 nd millennium BC, and the accompanying processes of 
assimilation and integration that define the origins of ethnic groups and the 
whole process of ethnogenesis not only in the steppes but also in neighboring 
regions of the Old World that may provide clues to the solution of the Indo- 
Iranian homeland problem. 

The approach to ethnic reconstruction utilized in this book draws from a 
number of different sources. 

1. Written evidence and linguistic data concerning the material culture of the 
ancient Indo-Iranians 

The evidence of the Indo-Iranian languages, including toponyms and onomastics 
as well as the earliest written tradition — the Iranian Avesta and the major Indian 
Sanskrit texts of the Rigveda, Atharvaveda, Yajurveda, and Satapatha- 
Brahmana — provide a basis for reconstructing Indo-Iranian material culture. The 
use of these texts is rather complicated as they were compiled, in the first place, 
after the Indo-Iranians had left their homeland and they do not contain 
contemporary cultural descriptions but only reminiscences of the homeland and 
ancestral culture obscured by their mythological context. Secondly, the creation 
of the texts themselves was separated by several or more centuries from their 
surviving written form, which has led to later interpolations, additions and 
lacunae. Thus, such sources require subtle analyses, a delicate handling of the 
information presented, and by no means do the sources permit a direct 
extrapolation of memories of the ancestral homeland culture carried in later 
Indo-Iranian tradition. General works on the history of India, Iran and Central 
Asia as well as works on specific categories of the material culture of the Indo- 
Iranians by W. Rau (1971; 1972; 1973; 1974; 1975; 1977; 1983), T. Ya. Eliza- 
renkova (1972; 1989; 1995; 1999b; Elizarenkova and Toporov 1995), F. B. J. 
Kuiper (1991), the important studies on the Avestan society by Zaehner (1961) 
and especially by V. I. Abaev (1956) and M. Boyce (1987 a,b; Boyce, Grenet 
1991) have been invaluable. 

In this work I have used the translations of the Rigveda and Atharvaveda of 
T. Ya. Elizarenkova; in some cases the editor has employed other translations, 
especially those of W. Doniger O' Flaherty. The Avestan texts have been 
reproduced in translations made by S. N. Sokolov, V. A. Livshits, V. I. Abaev, I. 
M. Steblin-Kamensky, and I. S. Braginsky. I have also availed myself of the 
translations of J. Darmesteter. Other data employed in this work have been 



INTRODUCTION XV11 

drawn from later sources such as the Mahdbharata, Shahname, Narts (the epics 
of the Ossetes), and evidence from classical authors. 

2. Archaeological sources for studying the material culture of the Indo-Iranians 
The Indo-Iranian material culture reconstructed on the basis of linguistic data 
and written tradition must then be compared with the material culture of two 
zones of the Old World — southwest Asia/India and the Eurasian steppe — in 
order to establish the Indo-Iranian homeland. Special attention is paid to the 
Andronovo culture, the creators of which are viewed by many linguists and 
archaeologists as the earliest Indo-Iranians. Comparative analysis permits us to 
verify the attribution of the Indo-Iranians to the Andronovo culture. However, as 
the Andronovo culture has never been the subject of a monograph, a preliminary 
collection and systematization of Andronovo materials was necessary. 

3. Archaeological sources on the material culture of Iranian-speaking Sar- 
matians and Saka 

These sources are thoroughly reviewed in the works by A. A. lessen (1953), S. I. 
Rudenko (1953), K. A. Akishev and G. A. Kushaev (1963), K. A. Akishev 
(1973), A. I. Melyukova (1964; ed. 1989), K. F. Smirnov (1964), M. K. 
Kadyrbaev (1966a); B. N. Grakov (1971; 1977), B. A. Litvinsky (1972), V. A. 
Il'inskaya and A. I. Terenozhkin (1983), M. G. Moshkova (ed. 1992), L. T. 
Yablonsky (1996), M. A. Itina and L. T. Yablonsky (1997; 2001), A. Yu. 
Alekseev (2003), etc. and they constitute the basis of the present research. 

4. Ethnographic sources for studying the material culture of the Iranian and 
Indian peoples 

The author is acquainted with ethnographic sources not only from the literature 
but also by taking part in ethnographic expeditions to the Pamirs in the 1950s, to 
Ossetia in the 1970s, and to Iran, Afghanistan, India, Sri-Lanka in the 1980s, 
where long isolation has preserved some rather archaic cultural features of the 
Indo- Aryans. 

Research on the peoples of northwest Hindustan was of special interest 
because there, in conditions of isolation, we find different Indo-Iranian tribes 
which, according to G. Morgenstierne (1973), left the homeland early, settled in 
the mountainous valleys and avoided the influence of the Harappan culture. They 
have preserved archaic features in language, mythology, folklore, way of life and 
handicrafts (Jettmar 1975; Fussman 1977; Rye and Evans 1976). 

5. Anthropological sources 

To verify the ethnogenetic conclusions, the results of anthropological research 
have also been applied. These have included the general works of V. P. Alekseev 
(1981; 1986; 1989; 1990), V. P. Alekseev and I. I. Gokhman (1984), T. K. Kho- 
dzhayov (1983), and V. A. Dremov (1997). 



A cknowledgements 

This book is a completely revised edition of the monograph "Where do Indo- 
Aryans come from?", which was completed in 1992 and published in 1994 in 



XV111 INTRODUCTION 

Russian. The book presents a summary of the Andronovo pastoralists of the 2nd 
millennium BC, whose representatives have been regarded by many scholars as 
the earliest Indo-Iranians. This hypothesis has been demonstrated on the basis of 
a detailed comparison of archaeological data and Indo-Iranian sources. 

Much new material has recently emerged in connection with the origins of 
the Andronovo culture, its metallurgy development, and its use of horse-drawn 
chariots. There is also new information about the migration of early Andronovo 
tribes to the south, into Central Asia, and their contacts with the Anau, Oxus, and 
Sarazm cultures, whose population belonged to the early farming cultures of the 
Near East. 

The use of calibrated radiocarbon data raises the question of setting the 
beginning of the Andronovo culture to the borderline of 3 rd / 2 nd millennium BC. 

The book has been considerably recast, new paragraphs and material have 
been added, because I considered it necessary to include the new data. However, 
some details and particulars, which may not be of interest for a foreign reader, 
have been excluded. As a result, this edition has been considerably extended. 

The scientific edition of this book would have been impossible without Prof. 
J. P. Mallory, who bravely volunteered to perform this huge task. I have been 
inspired by his priceless, useful, productive, and stimulating ideas on the genesis 
of Indo-Europeans. 

I would also like to express my heartfelt gratitude to Prof. Alexander 
Lubotsky, who suggested publishing my book in his prestigious series. 

I am also grateful to Proff. Vladimir Livshits and Tatyana Elizarenkova who 
kindly checked the transliteration of Indo-Iranian words used in the book. 

I am very much pleased to express my gratitude to V. V. Ivanov, my guru 
during many years, whom I respect most deeply in spite of the difference of our 
conceptions. 

A. M. Mandel' shtam and M. P. Gryaznov, the first readers of my book, 
should be also mentioned with gratitude, as well as M. N. Pogrebova, the editor 
of the first edition. Their advice and, what is more important, moral support in a 
difficult period of my life have been very important for me. 

The publication of this book has been realized after Catheryn Linduff and 
Karin Rubinson generously offered financial support. I express my gratitude to 
them. I am equally thankful to Peabody Museum, Harvard University, and Prof. 
Carl Lamberg-Karlovsky for their financial aid for the translation. I also want to 
express my gratitude to the director of Institute for Cultural Research, Kirill 
Raslogov, who found the possibility to sponsor my work in the period of a 
reduction of funding for the institute. 

I would like to express my sincere thanks to my son Alex Kuz'min for his 
constant financial and moral support, invaluable advice, patience, and courage. 

Dr. Natalia Krikova and especially my pupil Dr. Aleksandra Ippolitova 
performed a lot of technical work preparing the text. All these people graciously 
helped me, and my heart-felt thanks go to them. 

Finally, I would also like to mention all my colleagues, museum and library 
staff all over the world who helped me work with new materials and books on 
archaeology. 

Dr. Svetlana Pitina and P. Prudovsky have kindly translated the book. 



PART ONE 
THE ANDRONOVO CULTURAL ENTITY 



CHAPTER ONE 
HISTORY OF RESEARCH ON THE ANDRONOVO CULTURE 

Andronovo sites extend over 3,000km across the steppe and forest-steppe, from 
the western slopes of the Ural mountains as far east as the Yenisey; in the north 
the Andronovo culture reached the borders of the taiga zone and in the south it 
extended up into the high mountainous regions of the Tian-Shan and the Pamirs 
and penetrated the deserts and oases of Central Asia as far as the left bank of the 
Amu -Darya (Oxus). 

One can discern four stages in the development of Andronovo studies: 1) Pre- 
revolutionary; 2) 1920s - 1947; 3) 1947 - beginning of the 1970s; 4) Modern. 

In 1825 S. B. Bronevsky uncovered individual burials that were later 
assigned to the Andronovo culture of Kazakhstan; in the 1850s A. N. Zyryanov 
was opening burials in the Urals; in 1862 V. V. Radlov conducted the first scien- 
tific excavations near Karkaralinsk. At the end of the 19th / beginning of the 20th 
century, Andronovo graves were being studied in western Siberia, Kazakhstan, 
and in the Urals. One should especially note the works of N. K. Minko, V. Y. 
Tolmachev, and Yu. P. Argentovsky. Collections of metal objects were made, 
including a large one assembled by the Beloslyudovs. 

The second stage of research dates from the 1920s to 1947. In S. A. 
Teploukhov's report to the Russian Museum the Andronovo culture was singled 
out as a special culture of the Minusinsk region and was named after a cemetery 
near the village of Andronovo near Achinsk, which was studied in 1914 by A. 
Ya. Tugarinov (Teploukhov 1929a: 43-44, 58; 1929b). In 1929 M. P. Gryaznov 
outlined for the first time the chronological and territorial borders of the 
Andronovo culture, situating it in western Siberia, the Urals and Kazakhstan, and 
he synchronized the culture with the Timber-grave and Seyma periods. 

The establishment of Marxist methodology in archaeology directed research 
toward the study of the productive forces of the peoples of the USSR who had no 
written history. This led to the organization of expeditions to Kazakhstan, 
Siberia, the Tian-Shan, and Central Asia, and the excavation of settlements, 
houses and industrial complexes which were extremely important for the study 
of the history of production. This form of economic archaeology became the 
most important contribution of Soviet scientists to the development of archaeo- 
logical thought. Excavations of the Alekseevka settlement and the cemetery by 
O. A. Krivtsova-Grakova (1948) must be especially mentioned, as she investiga- 
ted the economy, way of life and ideology of an early group of people. Research 
into ancient mining and the settlements of metallurgists in Kazakhstan, under- 
taken by S. S. Chernikov (1939, 1948, 1949), made it possible to discuss for the 
first time the question of the organization and technology of the metal industry. 
This was accomplished at a time when much debate revolved around outdated 
models of migration (where Gustav Kossina's approach had been adopted by the 



4 CHAPTER ONE 

Nazis) and N. Ya. Marr's theories strengthened the ideas of autochthonous 
(stadial) development. Both of these trends downgraded research into ethno- 
genetic questions. 

The year 1947 saw the beginning of the third stage. The 'First Urals Archaeo- 
logical Conference' and publication of the monograph by O. A. Krivtsova- 
Grakova The Alekseevka Settlement and Cemetery and a number of works by S. 
S. Chernikov (1948, 1949) on the history of mining were dedicated to studying 
ancient economic systems. The work by S. V. Kiselev The Ancient History of 
Southern Siberia (1949) viewed the Andronovo culture for the first time against 
a wider background of synchronous cultures of the Old World, summed up pre- 
war Andronovo culture studies, and was a major event in Andronovo research. S. 
V. Kiselev tackled questions on the origin, cultural relationships, historical fate, 
specific character of the economy, role of farming, and social differentiation in 
the Andronovo culture. Generalizing works by S. P. Tolstov (1948) and A. N. 
Bernshtam (1949; 1950; 1952) were also published. They revealed the centuries- 
old development of culture in Central Asia involving it in the world historical 
process. These pioneer works were highly appreciated by the leader of European 
archaeology, V. Gordon Childe, who contributed so much to establishing the 
historical method in archaeology. 

The 'First Urals Archaeological Conference' raised questions on the 
classification and realization of regional and chronological differences of various 
Andronovo sites. K. V. Sal'nikov (1951) suggested his own division of the 
Trans-Urals Andronovo culture and proposed three stages: 1) Fedorovo (15th- 
13th centuries BC); 2) the genetically connected Alakul' (11 th 9th centuries 
BC); 3) Zamaraevo (8th-7th centuries BC), which provided the basis for the 
culture of the Iranian-speaking Scythian -Sauromatian tribes. 

V. N. Chernetsov opposed K. V. Sal'nikov's scheme (1953; 1963). Expert in 
Finno-Ugrian folklore and ethnography, a specialist on the forest cultures of the 
Urals and western Siberia, he suggested a genetic connection between the Ugrian 
population of the Trans Urals and the Andronovo culture which he held to be the 
descendants of the local Neolithic population related to the Kelteminar culture; 
his evidence was the similarity between the ornament on Andronovo and Ugrian 
material and their closeness with earlier Neolithic ceramics. He proposed 
dividing the Andronovo culture into two genetically unrelated cultures: the Fedo- 
rovo, related to the forest cultures of the Trans-Urals, and the Alakul', connected 
with the southern steppe regions and the Timber-grave culture of the Volga. 

A special role in the debate was played by O. A. Krivtsova-Grakova (1948: 
149-153, 161). Starting from the well-developed classification of the Timber- 
grave culture she outlined two stages: a late Alekseevka, dating from the end of 
the 2nd / beginning of the 1st millennium BC, and an early Kozhumberdy, syn- 
chronous with the Fedorovo in Siberia, where its evolution was interrupted by 
the intrusion of the Karasuk tribes. 

Thus in 1948 three different points of view on the chronology and historical 
fate of the Andronovo culture took shape. All subsequent researchers studying 
Andronovo problems adhered to one or other of the earlier hypotheses. 

Taking into account the heterogeneity as well as the continuity of Andronovo 
sites, A. A. Formozov (1951b) introduced the notion of the "Andronovo cultural 
unity". 



HISTORY OF RESEARCH ON THE ANDRONOVO CULTURE 5 

The 1950s and 1960s were marked by the rapid development of field studies 
of the Andronovo culture. The accumulation of a wide body of new material 
made it possible to divide it into separate local variants. In 1966 the Kazakhstan 
sites were classified according to K. V. Sal'nikov's scheme: Fedorovo (Nura), 
Alakul' (Atasu) stages; sites of the Final Bronze Age were designated the 
Begazy-Dandybay culture (Margulan et al. 1966). A. Kh. Margulan's mono- 
graph (1979) was extremely important for resolving problems of the Final 
Bronze Age as it systematized material from the Begazy-Dandybay culture 
(12th-9th centuries BC). 

The Andronovo culture of eastern Kazakhstan was studied by S. S. 
Chernikov (1960). He divided it into four chronological stages: 1) Ust'-Bukon, 
where ceramics possessed some affinities with the Afanas'evo and Okunevo 
cultures; 2) Kanay (16th— 12th centuries BC); 3) Malokrasnoyarka (1 1th 10th 
centuries BC); and 4) Trushnikovo (9th-8th centuries BC). 

M. N. Komarova (1962), following K. V. Sal'nikov's scheme, divided all 
Andronovo sites into Fedorovo and Alakul' types, and distinguished local 
variants. She maintained that only ceramics of the Fedorovo type were present in 
western Siberia, and only ceramics of the Alakul' type were represented in 
western Kazakhstan and Kustanay. 

The works by S. V. Kiselev (1960; 1962) and N. Ya. Merpert (1962) on 
changes in chronology, and the earlier dating of the Bronze Age cultures of the 
Eurasian steppes based on a wide ranging chronology — from China to Western 
Europe — were extremely important for resolving problems of the chronological 
position of the Andronovo culture. The establishment of a longer chronology 
demanded corrections to be made to the Andronovo schemes as well. The 
chronology of the Alakul' phase was extended because of new discoveries of 
metal objects in the Bliznetsy (15th 12th centuries BC) and Emba burials 
(Kuz'mina 1961b: 90-93, fig. 32.4.6.10; 1963a: 136). 

Because of the discovery at Bliznetsy of an Alakul' vessel typologically close 
to that of the Fedorovo phase, E. F. Fedorova-Davydova (1960; 1964) accepted 
V. N. Chernetsov's hypothesis and acknowledged that the Alakul' and Fedorovo 
types were synchronous local cultures. These ideas were supported by M. F. 
Kosarev (1965: 243). 

To improve the chronology I set about to study the Elenovka micro-region in 
the southern Urals. Applying for the first time aerial-photography fifty settle- 
ments, adjoining burial grounds, and a number of mines were discovered. Their 
chronology has been established on the basis of the first statistical approach 
applied to the classification of Andronovo ceramics. It is possible to discern 
three types of settlements and corresponding burial grounds and transfer the 
Elenovka model to other sites of the Orsk variant. 

The study of sites in a micro-region, applying the latest methods of research 
— aerial-photography, dendrochronology, the spectral analysis of metals, 
compositional analysis of ceramics and the investigation of adjacent mines — also 
made it possible to pronounce on the issues of Andronovo demography, 
exchange, and domestic architecture (Kuz'mina 1962a; 1963b; 1963g: 133-138; 
1965a). 

The works by E. N. Chernykh (1970), who singled out an Elenovka-Ushkatta 
group among the metal types of the Eurasian steppes, were important for the 



6 CHAPTER ONE 

study of Andronovo metallurgy. V. I. Tsalkin (1972b) investigated the charac- 
teristics of Andronovo stock-breeding. 

Great attention to the Andronovo culture was raised by K. V. Sal'nikov 
(1967: 315-325) in a generalizing work discussing questions of the development 
and interrelationship of the cultures of the Urals. He dated the Fedorovo stage to 
the 18th-16th centuries BC, the Alakul' to the 15th-12th centuries BC, and the 
Zamaraevo to the 12th— 8th centuries BC. 

V. S. Stokolos sharply criticized K. V. Sal'nikov's scheme (1967; 1972). He 
concluded that the Alakul' monuments are more ancient than the Fedorovo and 
belonged to a different culture. The work by V. S. Stokolos was in turn subjected 
to sharp criticism due to numerous mistakes. Although considerable progress had 
been made, the 1960s did not result in a single conception of the Andronovo 
culture. 

The fourth stage of Andronovo studies has been characterized by more active 
research on the periphery. Successive surveys have been undertaken in central 
Kazakhstan (Kadyrbaev 1969; 1972; 1974; 1983), especially in the Karaganda 
region (Evdokimov 1979; 1980; 1983; 1984; 1987). Intensive research was also 
carried out by an expedition from Sverdlovsk University. G. B. Zdanovich 
(1975, 1988) published his research on the north Kazakhstan expedition. He 
studied multi-layer sites, reliably established their vertical stratigraphy, and iden- 
tified the Petrovka type of sites, preceding and genetically connected with the 
Alakul' ones. The Petrovka type is dated to the 15th century BC, the Alakul' to 
the 14th 13th centuries BC, the Fedorovo and mixed Amangel'dy to the 12th- 
11th centuries BC. 

The scheme for the Final Bronze Age of the northern Kazakhstan Bronze 
Age was formulated by S. Ya. Zdanovich (1974b; 1979), who defined a separate 
Sargary culture (other scholars assigned the sites to the Alekseevka type), dating 
it to the 9th-7th centuries BC. The Zdanovichs' scheme was an important 
contribution to the study of the Andronovo culture but their absolute chronology 
required verification. 

The discovery of early burial complexes in 1973 in the Urals in kurgan 25 of 
the Novokumak cemetery by K. F. Smirnov (1973: 175-176) and in the Sintashta 
cemetery by V. F. Gening and L. I. Ashikhmina (1973: 132-133; Gening 1975: 
144-147; Gening 1975a: 94; Gening, 1977) opened a new page in the study of 
the Andronovo culture. The examination of the Novokumak burial assemblage, 
based on a wide comparative and historical background, led to the definition of a 
chronological horizon comprising the Sintashta and Petrovka type monuments of 
the Urals and western and northern Kazakhstan. This stage is dated to the 17th- 
16th centuries BC on the basis of the discoveries of metal objects and bone disc- 
shaped cheek-pieces with analogies in the Ukraine, the Danube and Greece no 
later than the 16th century BC (Smirnov and Kuz'mina 1977). The formation of 
the Andronovo culture was viewed as a result of a western impulse and 
migration from Eastern Europe of the bearers of the Poltavka, Abashevo, and 
Multi-roller ceramic cultures (late Catacomb). Their assimilation led to the 
formation of the Andronovo culture, whose bearers were acknowledged as Indo- 
Iranians (Kuz'mina and Smirnov 1976; Smirnov and Kuz'mina 1977; Kuz'mina 
1981a). V. F. Gening came to the same conclusion (1977) while comparing the 
burial rite at Sintashta with Vedic data. 



HISTORY OF RESEARCH ON THE ANDRONOVO CULTURE 7 

N. A. Avanesova (1979) suggested a periodization of the Andronovo culture 
on the basis of the typology of metal objects. She accepted the unitary nature of 
the Andronovo culture and regarded it impossible to divide it into independent 
Fedorovo and Alakul' cultures. She suggested a pre-Alakul' stage (17th— 16th 
centuries BC); Alakul' (15th— 14th centuries BC); Kozhumberdy (14th century 
BC); Fedorovo (end of 14th 13th centuries BC); and Zamaraevo-Begazy culture 
(12th-9th centuries BC). From my point of view, ignoring the burial rite and 
ceramics in her analysis led N. A. Avanesova to a number of erroneous 
conclusions. 

M. P. Gryaznov (1969) and G. A. Maksimenkov (1978) presented important 
summaries of the critical evidence from Siberia that bears on the Andronovo 
problem. G. A. Maksimenkov stresses the unity of the Andronovo culture, and 
the early date of the Fedorovo sites presupposes an arrival of Andronovo tribes 
in Siberia from the west. 

Important discoveries leading in many cases to a new understanding of the 
ancient history of Siberia are connected with the research of V. I. Matyush- 
chenko (1961; 1969a; 1969b; 1970; 1973a; 1973b; 1975; 1978; Matyushchenko 
and Lozhnikova 1969; Matyushchenko and Sinitsina 1988). He discovered 
spectacular cultural material from the settlements at Samus' and Rostovka, and 
the Elovka cemetery, with its richly represented metal objects; these discoveries 
raised the question of contacts with the pre -taiga zone. 

Problems concerning the cultural interaction between hunter and fisher tribes 
with southern pastoralists are the focus of M. F. Kosarev's works (1965; 1970; 
1974; 1981; 1983), which traced the influence of the Andronovo culture on the 
forest zone. 

New material bearing on these questions and the problems of the 
ethnogenesis of the peoples of southern Siberia arose from the research on the 
Preobrazhenka 3 and Sopka 2 complexes (Molodin 1973; 1974; 1975; 1977; 
1979; 1981; 1983a; 1983b; 1985), and Fedorovo-type sites in the Altai, whose 
creators Yu. F. Kiryushin (1980; 1983) considered to be Iranians. A. V. Matveev 
(1995) carefully investigated the chronological problems, and placed the 
Novokumak horizon between the 22nd and 18th centuries BC according to the 
evidence of radiocarbon dating. 

Specialists studying Siberia with their historical method of study, their wide 
diachronic approach and focus on ethnogenetic issues universally recognize the 
alien character of the Fedorovo tribes in Siberia and the absorption of their cul- 
ture by local tribes under powerful Andronovo influence. Considerable success 
has been achieved in studying the so-called Andronovoid 1 cultures of the taiga 
natives under Andronovo influence (Obydenov and Shorin 1995; Obydenov 
1997; Potemkina et al. 1995). 

Of special value are those studies that have concentrated on the relationship 
between the Andronovo culture and the cultures of western Central Asia. Sites 
combining Andronovo ceramics and metal with vessels of the Bishkent (Vakhsh) 
and farming culture of the Namazga VI stage have been recorded for Turkmenia 
Bactria and Margiana (Mandel'shtam 1960; Litvinsky 1964; Kuz'mina 1963 
1980; 1984; 1988; 1990; 1993; 1994; Vinogradova and Kuz'mina 1986; 1996 



Editor's note: clumsy but probably better English than Russian 'Andronoid'. 



CHAPTER ONE 



Francfort 1981; 1989; Lyonnet, 1996; P'yankova 1986; 1990; 1998; Avanesova 
1996). 



CHAPTER TWO 

METHODOLOGICAL ASPECTS OF ETHNOCULTURAL 
RECONSTRUCTION 

Archaeological taxonomy comprises a number of key concepts for the 
classification of archaeological material. These include 'archaeological culture' 
and 'archaeological cultural unity' which relate to such basic notions of theo- 
retical ethnology as 'economic and cultural type' and 'historical and ethno- 
graphic unity'. 

The initial archaeological category is the notion of the 'artifact', i.e. "any 
material object produced or modified by man in the past performing a certain 
cultural function" (Bochkarev 1975: 35). The totality of artifacts comprises a 
complex, i.e., the empirical basis of archaeology. 

The first level of interpretation of archaeological material will be the 
classification of artifacts according to types. Oscar Montelius introduced the 
theoretical basis of archaeological classification, commonly understood as the 
typological method. However, there is still no unity of understanding as to what 
precisely an archaeological 'type' means. Some view the type as an actual 
existing and historically conditioned phenomenon (Sher 1966; Gryaznov 1969; 
Kamenetsky 1970); others (Ford 1962; Fedorov-Davydov 1970) understand this 
category as purely gnoseological, a purely subjective instrument of research. It 
seems that, as a rule, the artifact type was understood by its creator as a real 
entity within a culture. This is proved by differences in settlement inventory and 
related cemeteries (the presence of ritualized and non-ritualized artifacts), and by 
the fact that archaeologically distinguished artifact types have independent 
names in the language of a culture. 

Typological classification consists of the isolation of peculiarities of artifacts, 
but it also takes into account the interconnection of characteristic features. This 
permits one to reconstruct historically existing types within a culture and not a 
priori classes of artifacts. 

The second level of interpretation of archaeological materials is connected 
with the typological classification of complexes, i.e., with distinguishing the 
types of sites according to Yu. N. Zakharuk (1981: 20-21). The site type is "a 
universal unit of classification both for defining within some small territory 
genetic connections and for discerning synchronous groups." Systematization at 
this level is built on the correlation of the elements of burial rite with the basic 
categories of material culture, primarily ceramics. Then domestic settlement 
wares are compared with mortuary ceramics. Site types outlined according to a 
set combination of principles are then mapped. Their relative chronology is 
established on the basis of stratigraphical data, transparent typological series and 
mutually associated types within closed complexes. The absolute age of every 
site is defined on the basis of imports and such chronological indicators as bone 



10 CHAPTER TWO 

cheek-pieces and metal objects. Finally, according to the accepted definition of 
archaeological culture type, the chronological stages and local variants of the 
archaeological culture are evaluated and the degree of similarity between sites 
and with foreign cultures is determined. 

The third level of archaeological analysis consists of distinguishing 
archaeological cultures or an archaeological cultural unity. This involves a 
transition from strictly archaeological research to historical interpretation. The 
archaeological culture is an open, dynamic, statistically stable system of different 
types of sites that occupy a continuous (though sometimes changing through 
time) territory with an objectively established unity of interconnected types, 
which develop uniformly over a long time period and vary in space in a limited 
manner, distinguishing this system markedly from other systems (i.e., 
archaeological cultures) (Kamenetsky 1970; Kovalevskaya 1988; Kuz'mina 
1988d). 

Sometimes divergence in the territorial aspect is such that some 
environmental zones exhibit independent lines of development. They do not 
coincide in everything; quite often sites of different types are situated like strip- 
farming. 

In some cases where there is no satisfactory criteria for defining an 
archaeological culture according to the strict definition of recurring principles 
that occur uniformly in time and within a restricted area we may still discern a 
chain or lattice of interconnected site types that sharply differ from foreign ones. 
In these cases there is also reason to unite these sites into a cultural and historical 
unity. Sites of these archaeological 'unities' may have constrained time-space 
limits and are united according to one but differentiated according to other fea- 
tures. What is especially important is that they can be sharply distinguished from 
other sites. The notion of the archaeological cultural unity (or cultural region) 
was first introduced by A. A. Formozov (1951b; 1959b) and N. Ya. Merpert 
(1968; 1985) with respect to the Andronovo, Pit- and Timber -grave sites. 

What is the historical reality of an archaeological culture or cultural unity? 
Some ethnologists and archaeologists are sceptical of identifying the 
archaeological culture with the ethnos (Foss 1952; Arutyunov and Khazanov 
1978; Klein 1978, Clarke 1978; Renfrew 1977). Others see the archaeological 
culture as a reflection of an ethnos in terms of a social unit (Gorodtsov 1927; 
Bryusov 1956; Formozov 1959a; Zakharuk 1964; 1981; Artamonov 1969; Grya- 
znov 1969; Tretyakov 1969; Gening 1970; 1976; Kamenetsky 1970; Kuz'mina 
1976a; 1988d; Bromley 1983; Merpert 1985; Alekseev 1986; 1989; Kovalevs- 
kaya 1988). Of decisive importance for resolving the problem of the relationship 
between archaeological culture and ethnos is the concept of cultural tradition. S. 
A. Arutyunov (1989) demonstrated that a cultural tradition is revealed first of all 
in language and meta-cultural communication (art, family ritual and burial rite). 
'Cultural tradition' is a stable combination of an interrelated and uniform system 
of elements of material and spiritual culture as well as the very ways of 
transferring information from generation to generation. This system is capable of 
preserving and sustaining itself with a high degree of regularity through stages of 
cultural development, retaining its stability over a long period of time in order to 
reproduce an ethnos in a certain ecological niche. 



METHODOLOGICAL ASPECTS OF ETHNOCULTURAL RECONSTRUCTION 1 1 

Traditional culture includes the type of settlement, architecture, tools, 
everyday objects, dress, attributes of cultural activity, works of art, techniques of 
production, especially functionally irrelevant methods of production and artistic 
styles that are specific to a given ethnic group and serve as certain ethnic 
indicators. It also includes norms of behavior, etiquette, dance, certain 
ornamental motifs and pictorial art, images and compositions of decorative art 
and the whole complex of spiritual culture, including mythological ideas, epics, 
and other forms of folklore in more developed societies. 

Traditional culture comprises an integral complex where all things and 
actions carry a semantic load connected with the mythological system and the 
self-consciousness of a given ethnic group; this contrasts with modern culture 
which distinguishes between the material and spiritual. Ethnically important 
axiological directives and ideological and aesthetic ideas provide an ethnic 
tradition. 

Cultural tradition is transferred to the next generation by instructing it in the 
form of ritual actions, etiquette, dances, and the artistic performances and images 
of mythological and epic narrative. Language is the main channel of commu- 
nication (Arutyunov 1989; Arutyunov and Cheboksarov 1972). A common 
language is not only the main sign but also the main condition for forming and 
preserving a traditional culture. The unity of a language within a culture is, as a 
rule, ideologically fixed in a common native name and self-consciousness, which 
is a characteristic sign of the ethnos and one of the conditions of its reproduction. 

In cases where there is ethnic and linguistic divergence involving the 
dispersion of a group into different territories, the notions of a former kinship of 
the separated groups are firmly fixed in the ideological sphere in the form of 
legends about a common primary ancestor and the familial relationship between 
separate tribes. But in cases involving the amalgamation of two ethnic groups 
into a new culture, ethnographers have shown that after a bilingual period, one 
language dominates and a mythologized notion of a common origin is created. 
Thus, an archaeological culture is a direct reflection of a living ethnographic 
culture and it must also be characterized by the linguistic unity of its 
representatives. 

The archaeological culture or cultural unity meets the requirement of an 
historical and ethnographic region (Levin and Cheboksarov 1955) and it 
represents a certain stage of the development of the ethnos or meta-ethnos (Bruk 
and Cheboksarov 1976; Bromley 1983). Research supports the correspondence 
between actually existing ethnoses, attested from written sources, and 
archaeological cultures. 

A. A. Formozov (1959b) posed the question of the correlation between the 
archaeological cultural and an actual ethnic and social unity. He introduced the 
notion of ethnic and cultural unity that meets the requirements of an historical 
and ethnographic region according to Levin and Cheboksarov. 

So an archaeological cultural unity is a dynamic system of closely related 
(due to either a common origin or the process of the consolidation of different 
ethnoses) interconnected sub-ethnoses, forming a continuous chain or network 
within a limited natural habitat (probably changing in time). In this network there 
is a coincidence of the majority of ethnic markers. This does not exclude 
possible local and temporal variability within this unity, including differences in 



12 CHAPTER TWO 

the history of the most distantly separated sub-ethnoses in the network, but it 
does presuppose a unity of the main categories of culture and their membership 
in a single language family. Tribes constituting a cultural unity may speak 
different dialects of one language or already separated languages of a single 
language family, thus maintaining linguistic continuity. 

Taking into account the fact that the site type is defined by the totality of a 
cultural unity, and that sites are territorially grouped, the sub-ethnos of a tribe or 
group of closely related clans can be assumed. At the same time the site type is 
the main structural unit of an archaeological culture which one naturally likens to 
an ethnos (such as a tribe). The site type is also the basic unit of the 
archaeological cultural unity, but the size of the natural environment and 
considerably greater variety within the unity directs us to compare the 
archaeological cultural unity to a meta-ethnos and not an ethnos. 

Economic activity and the corresponding economic structure condition the 
existence of the ethnos. Following Leslie White (1959) I acknowledge the 
considerable influence of ecological factors on the development of an ethnic 
culture (Kuz'mina 1996). The basis of White's thesis is the concept of culture as 
the extra-somatic adaptation of human groups to environmental influence. He 
acknowledges that necessity to overcome crisis stimulates cultural innovation. At 
the same time a great role is attributed to the increasing thermal or pyrotechnic 
potential of society, which is viewed as a major factor in human history. 

In the context of this work the definition of ethnic culture is based on the 
approaches developed by E. S. Markaryan (1973a; 1973b), S. A. Arutyunov 
(1989), and V. P. Alekseev (1993). Along with White they see ethnic culture as a 
combination of non-biologically devised mechanisms that are interconnected and 
transferred from one generation to another to provide an adjustment to a certain 
ecological niche in order to reproduce the ethnic group. 

Economic development is defined by a number of factors: the productivity of 
the environment, the presence of minerals, natural forces, including climatic 
changes and humidity, anthropogenic factors such as the destruction of natural 
resources, soil exhaustion, salinization, and desertification of soils as a result of 
extensive farming and cattle-raising, and, finally, population growth that exceeds 
the resources of the environment. 

Progress in production undergoes a basic evolutionary process. Innovations 
are present in any traditional society, but their number is not large, and their 
dissemination is slow. But in periods where many unfavorable factors conspire 
together, this leads to a crisis of the traditional economy. The crisis might result 
in the entire abandonment of a territory by the whole population or a part of it 
through migration, or a major shift in the form of the economy. 

Crises forced the search for solutions, they stimulated invention, created 
conditions for intellectual development and enhanced the opportunity for 
innovation. Innovations and an increase in productivity provided the surpluses 
that allowed for more time for intellectual activity. Cultural innovations were 
ideologically incorporated by being given roots in the mythological system and 
were depicted in art, all of which led to a fundamental transformation of the 
whole cultural image. That is why it would be unjustified to speak in terms only 
of social and economic progress in some periods rather than the intellectual and 



METHODOLOGICAL ASPECTS OF ETHNOCULTURAL RECONSTRUCTION 1 3 

economic development of society. The defining role of ideology and intellectual 
activity for the development of society has not been sufficiently evaluated. 

The concept of economic and cultural type (ECT) was introduced by M . G. 
Levin and N. N. Cheboksarov (1955) and developed further (Andrianov 1968; 
Andrianov and Cheboksarov 1972; Chesnov 1970; 1982; Vaynshtein 1973, 
Bromley 1973; Alekseev 1979). It is connected with the problem of the 
adaptation of the ethnos to the environment. The adaptation to a similar 
ecological environment conditions a similar character of economic activity and 
the formation of similar features of material culture for different ethnoses, 
including economy, foodstuffs, tool types, domestic architecture, and costume. 
The culture complex comprises two main groups: practically all the components 
of the ECT form the first group while specific features of culture that are not 
directly related to the economy and adaptation to the type of environment 
constitute the second group. The analysis of parallel economic and cultural types 
and the evolution of the ethnos reveals that the ECT, at least in ancient times, has 
important ethno-differentiating and ethno-integrating properties (Chesnov 1982). 
Some signs characterizing the ECT are ethnically specific, although they 
originate under the influence of ecological factors — foodstuffs, house type, 
dress, industrial technology of the various components of material culture, 
especially ritualized. They can be preserved in the culture of the ethnos for a 
long period of time, even over the course of migration, change of environment 
and type of economy. In this case they become ethnic indicators (Kuz'mina 
1981a; 1986b). Ethno-diagnostic signs are peculiarities such as the form and 
ornamentation of ceramics, functionally unconditioned ways of production, and 
some types of weapons, details of burial ritual, etc. Only by viewing the 
interrelationship of these characteristics that reflect the systemic character of the 
archaeological culture (or cultural unity) as an ethno-social organism can one 
form a correct image of the archaeological culture and cultural unity. 

In order to define an archaeological culture or cultural unity it is required to 
accumulate the following evidence (Kuz'mina 1985a): 1) a complete collection 
of material within the entire natural territory to create both a resource base and to 
undertake a critical evaluation of the completeness and quality of the sources; 2) 
a complete list of all those traits that are required to describe the numerous sites 
of a territory and a typological classification of all sites according to a consistent 
combination of indicators; 3) cartographic analysis of all sites in question; 4) the 
establishment of a relative chronology of the sites on the basis of stratigraphy 
and the construction of a continuous typological series, and the establishment of 
an absolute chronology of every site by imports and synchronization with dated 
sites of other cultures using a single chronological scale as well as critically 
evaluated radio-carbon analysis; 5) an evaluation of the results by introducing 
criteria for establishing the degree of similarity between sites, stages, local vari- 
ants or separate cultures, i.e., a definition of site types within their chronological 
stages and/or local variants or ultimately separate cultures, each of which is 
described according to a set of consistent criteria that distinguishes the archaeo- 
logical cultures under review from others; and 6) verification of the results 
against newly discovered archaeological complexes within the culture area 
(cultural unity). 



14 CHAPTER TWO 

While the archaeological culture (cultural unity) corresponds to a certain 
ethnic and social entity, the first and main stage of historical interpretation of an 
archaeological culture (cultural unity) is paleo-economic reconstruction because 
economic and cultural type (ECT) is directly reflected in the archaeological 
record. 

The analysis of material culture reveals the dynamics of production not only 
through the study of the tools themselves but also the character of production. 
That is why technological principles are at the foundation of any suggested 
classificatory scheme. The technology of production governs to a considerable 
extent the specific form of an object. I have applied this method of correlating 
technology and object form in my classification of metal objects, bone cheek- 
pieces and ceramics (Kuz'mina 1966; 1980a; 1985a; 1986a; 1986b). 

The study of the technology of production and the implements themselves 
establishes the whole complex of material culture as an integral system that 
includes elements conditioned by an ECT, which has adjusted to a specific 
environment, as well as functionally unconditioned elements and production 
techniques that constitute the traditional culture. However, as indicated above, 
the archaeological culture (or cultural unity) reflects an ethnic and cultural unity 
that is defined not only by the ECT but by culture in general. In approaching 
culture as an integral system, there are two main subsystems: the material culture 
that is directly reflected in the archaeological record and the spiritual culture that 
includes social organization and ideology. 

The spiritual culture and some of its components can be recovered from 
archaeological materials only with the help of reconstruction. That is why the 
historical interpretation of an archaeological culture (unity) and its ethnic 
attribution can be based only on the analysis of material culture and an 
investigation of spiritual culture cannot be the initial stage in the creation of an 
ethnic hypothesis. 

As the decisive component of an ethnic culture is realized in language unity, 
ethnic attribution is impossible without systematic and complex usage of 
philological data alongside archaeological and ethnographic materials. To this 
may be added folk traditions of modern ethnic groups which convey the cultural 
traditions of their former linguistic groups. Anthropological data are also 
significant in ethno-genetic reconstruction as has been shown by V. P. Alekseev 
(1986). 

Generally, there are two methods of ethno-genetic reconstruction and ethnic 
attributions that may be employed. The first consists in demonstrating a 
continuous sequence and succession of archaeological cultures on a specific 
territory until the emergence of historically known ethnic groups. This method is 
a variant of a more general retrospective method, the leading role of which is 
rather widely accepted in ethno-genetic construction at present. The second 
method employs the evidence of functionally unconditioned ethnic markers 
whose connection with the economic and social dimensions of a society are so 
minimal that they preclude independent convergent origin in different ethnic 
groups (Formozov 1959b). Each of these methods has limitations and only a 
combination of the two makes ethno-genetic reconstruction or ethnic attribution 
sufficiently reliable (Kuz'mina 1981a; 1986c). 



METHODOLOGICAL ASPECTS OF ETHNOCULTURAL RECONSTRUCTION 1 5 

We have argued that there is no division in traditional culture between 
spiritual and material. Ceramic production, house building, textiles are seen as 
repeating in micro-space the act of creation employing specific rituals. Objects 
themselves possess certain semantics. The material culture outlined in this study 
is conditional and presupposes that only material culture can be objectively 
reconstructed from archaeological data in contradistinction to an existing 
ethnographic culture of an actual ethnos. The reconstruction of spiritual culture 
and social structure depends considerably on the ethnic attribution of the 
archaeological culture taken by the researcher. Different peoples vary in their 
interpretations of similar cultural practices. For example, a grave gift might be 
interpreted as payment for the ferryman Charon (Greeks) or as gifts for the 
ancestors (Chinese) or as gifts of the living to the dead (Africans). This is why 
the main components of spiritual culture can only be described in terms of 
archaeological materials in historical reconstruction. 

Employing the results of such reconstruction at the stage of creating a 
hypothesis of ethnic attribution, either through retrospective analysis or through 
the use of ethno -diagnostic criteria, is methodically incorrect as different ethnic 
traditions can interpret one and the same archaeological factor as different 
cultural phenomena. Thus, the reconstruction of spiritual culture largely depends 
on explicitly or implicitly assumed ethnic hypotheses. But coordinating the 
different aspects of reconstructed spiritual culture on the basis of ethnic 
attribution, generated from an analysis of material culture, can and must be one 
of the main instruments of verifying suggested ethnic attributes. 

The suggested methodology of ethnic attribution of an archaeological culture 
(unity) requires a systematic and complex approach to the following data: 1) 
language (including toponymies); 2) written tradition; 3) ethnography; and 4) 
anthropology. It can be reduced to the following: 

1. Establishing by the retrospective method the genetic connections of a 
culture with a subsequent one, whose ethnic identity is known from written 
sources. This is accomplished by constructing from each category of material 
culture a typological sequence of development of interrelated features that traces 
the evolution of those features that serve as ethnic indicators. 

2. The use of the combination method (employing the evidence of language 
and written sources) to establish by independent means the ascription of a certain 
ECT to an actual archaeological culture or cultures. 

3. The verification of the ethnic attribution of a culture (cultural unity) on the 
basis of migrations involving the culture (unity) in its natural habitat and 
adjoining regions. 

4. The verification of the ethnic attribution of an archaeological culture 
(unity) by anthropological materials. 

5. Testing the validity of an ethnic attribution by comparing it to data 
concerning putative ethnic contacts (groups of ethnoses) and bearers of other 
languages. 

6. Testing the validity of an ethnic attribution by superimposing an 
archaeological map on a map of toponyms and hydronyms. 

7. Testing the validity of an ethnic attribution by assessing the spiritual 
culture reconstructed from archaeological data (burial rite, rituals, images and 



16 CHAPTER TWO 

compositions of art, social structure) in the light of the suggested ethnic 
hypothesis. 

The problem of ethno-genetic reconstruction is closely connected with the 
ethnic attribution of archaeological cultures (cultural unities). In the first place, 
the true ethnic attribution of a series of archaeological cultures makes it possible 
to establish a sequence of ancestral cultures and to trace the main stages of ethnic 
history. The evidence of language and written tradition makes it possible to 
reconstruct the type of economy, social structure and ideological beliefs of the 
speakers of the given ancestral language (group of related languages); isoglosses 
and borrowed words can reveal contacts with other language groups. However, 
linguistic data do not reveal when and where the reconstructed phenomena took 
place. It is only by correlating linguistic data with the archaeological record that 
one can establish the chronology and ethnic history of a pre-literate period. Some 
key peculiarities of the reconstructed ECT, ideology and social structure help 
provide chronological indicators for reflected linguistic phenomena. 

To resolve the problem of the homeland and localization of migration routes 
it is necessary to clarify the correlations of the reconstructed ECT and the circle 
of archaeological cultures reconstructed on the basis of linguistic data. If we take 
the Indo-Iranian homeland as an example, then this methodology presupposes 
the following: 

1 . A definition of the cultural complex of the most ancient Indo-Iranians on 
the basis of linguistic data and tradition and the establishment of a correspon- 
dence between the ECT and actual archaeological cultures of a specific region. 

2. Establishment of the time and direction of any migrations. 

3. Analysis of the different migration types reflected in the archaeological 
data and in the different cultural and economic zones. 

The suggested methodology will be pursued as follows. The separate 
categories of Andronovo material culture will be reconstructed on the basis of 
data drawn from the whole Andronovo area. They will be shown to correspond 
to the Indo-Iranians as reconstructed from linguistic data and historical tradition 
reflected in written texts. They will then be assessed against two zones: Central 
Eurasia and the Indo-Near -Eastern. The succession of cultures of the Early Iron 
Age will be traced and ethnographic parallels will be revealed. In order to 
achieve objective conclusions the analysis of each category of material culture 
will be conducted independently and the results obtained will only be combined 
at the final stage of this attempt to locate an Indo-Iranian homeland. 



CHAPTER THREE 

CLASSIFICATION OF SITES AND THE PRIMARY FEATURES OF 
ANDRONOVO UNITY 

According to the methodology introduced in chapter 2 the author assembled an 
archaeological data bank. The work was conducted in the following sequence. 
The data bank was supplied with materials from 400 cemeteries and settlements 
of the entire area, of which about 100 sites derived from different regions in the 
Urals and Kazakhstan, were examined and of these, 25 were excavated by the 
author (Maps 2, 3). In addition, there are more than 100 sites from Central Asia 
(Map 4; 30 investigated and excavated with the participation of the author). 

A considerable part of the data is the result of the author's research, which 
began in 1958. 2 From 1959 the author led an expedition to the Elenovka group of 



2 Other excavated materials as well as hoards and chance finds were studied from the 
museum collections of the State Historical Museum, the State Museum of the East, the 
Museum of Anthropology of the Moscow State University, the Department of Prehistoric 
Cultures and department of Central Asia in the Hermitage and the Museum of 
Ethnography in St Petersburg. Other sources were historical and regional museums, 
archaeological laboratories of the various universities and pedagogical institutes in 
Orenburg, Orsk, Chelyabinsk, Troitsk, Syktyvkar, Perm', Izhevsk, Ufa, Nizhny Tagil, 
Ekaterinburg, Tyumen', Tobol'sk, Omsk, Barnaul, Kemerovo, Krasnoyarsk, Novosibirsk, 
and in the Institute of Archaeology and Ethnography of the Siberian Department of the 
Russian Academy of Sciences, in the Ethnological and Archaeological Museum of Tomsk 
University, in the museums of Aktyubinsk, Ural'sk and Emba, in the Central Museum of 
Kazakhstan in Almaty, in the Museum of Art and the Institute of Geology of the Kazakh- 
stan Academy of Sciences, in museums and archaeological laboratories at Karaganda, 
Petropavlovsk, Semipalatinsk, Ust'-Kamenogorsk, Dzhambul, Chimkent, and Bishkek, in 
the History and Art museums of Kirgizstan, the Institute of History and Archaeology of 
Hrgizstan, in Przheval'sk, in Tashkent, in the History and Art museums of Uzbekistan, 
the Department of the History of Geology of the Central Asian State University in Samar- 
kand, in the Museum of History of Uzbekistan, Afrasiab, the Institute of Archaeology of 
the Uzbekistan Academy of Sciences, in Bukhara, Nukus, Khiva, Termez, Fergana, in Du- 
shanbe in the museums of History and Ethnography, in the Institute of History, Archaeo- 
logy and Ethnography of Tadzhikistan, in Asgabat — in the museums of Turkmenistan 
History and Art, the Institute of History, Archaeology and Ethnography of Turkmenistan, 
and in the History Museum in Merv. Other collections consulted were the Ukrainian 
Historical Museum, the Ukrainian Institute of Archaeology in Kiev, the Institute of 
History, Archaeology and Ethnography of Tatarstan in Kazan', the pedagogical institutes, 
universities and regional museums in Samara, Saratov, Simbirsk, Volgograd, and finally 
some collections of school museums and private collections. (I find it a pleasant duty to 
express my deepest gratitude to all the museum workers who assisted my work and also to 
colleagues who regularly provided me with new unpublished materials). Outside the 
former Soviet Union, the author consulted the Siberian and Central Asian collections in 
the Helsinki Historical Museum, in the Louvre, Saint-Germain, and Chernuschi in Paris, 
the Metropolitan Museum in New York, in the British and Victoria and Albert museums 
in London, the Bode Museum in Berlin, the Turkish Historical Museum in Ankara and the 
National Museum in Istanbul. Certain work was conducted in the museums of Iran, 
Afghanistan, India and Sri-Lanka. Materials from the archives of the IA AS, LOIA, IHvIK, 



18 CHAPTER THREE 

the Orenburg region in its investigation of more than 50 sites in the Elenovka 
microdistrict of the southern Urals, including the Elenovka and Ushkatta mines 
and workshop complexes, settlements, and burials. 3 Once the material recovered 
from the Andronovo sites had been assembled, it was necessary to systematize 
the data according to criteria that would be applicable to the entire territory. 
Taking into account the classification of East European and Siberian cultures es- 
tablished by V. A. Gorodtsov and S. A. Teploukhov, the burial rite was deemed 
the most significant feature for the purposes of cultural and ethno-historical 
reconstruction. The following features were recorded: 1) type of surface 
construction; 2) type of grave pit and evidence of recutting; 3) orientation; 4) 
cremation or inhumation and posture of the deceased; 5) other ritual features; 6) 
grave goods; 7) animal sacrifice; and 8) funeral feast (Figs. 1; 57). 

The analysis of the technology of hand-made pottery largely comprises vessel 
form, technique of manufacture and decoration, and ornamental motifs; these are 
also very important ethnic indicators and are used as the basis for defining 
cultures, stages, local variants and types. The following analytical approaches 
have been employed in this work: 1) vessel manufacture and form according to 
the system of the author (Figs. 2, 11-13); 2) analysis of tempering/opening 
material; 3) surface treatment; 4) construction of ornament according to the 
system of S. V. Ivanov (1963) and S. V. Zotova (1965); 5) elements of ornament 
after M. N. Komarova (1962); 6) organization of decorative elements within 
zones and their combinations; and 7) the technique of figuring ornamentation 
(Figs. 12; 13). 

The analysis was based on closed complexes of burials over the entire 
Andronovo area. These were classified within a single system, then the burial 
rite and ceramics were correlated, and then those sites with a stable assortment of 
features were united into types. The funerary ceramics were compared with the 
ceramics from the settlement and the latter were divided into types. We follow 
the definition of type employed by Yu. N. Zakharuk (1981) who regards it as a 
"universal unit of classification both for establishing within a small territory their 
genetic sequence over time and for establishing synchronous groups of sites on 
the territory of a single archaeological culture." Sites of every type were mapped 
and local variants were defined. 

IIA AS of Kazakhstan and a number of museums, personal archives of O. A. and B. N. 
Grakov, M. P. Gryaznov, S. S. Chernikov, V. S. Sorokin, A. N. Margulan were also 
employed. These sources were especially valuable in cases where the originals were now 
lost or exported. 

3 Excavations were made on some settlements (Ushkatta 1, 2, 7, 9, Kiimbay, Kupukhta, 
Baytu, Shandasha, Tursumbay, and cemeteries at Ushkatta, Ataken-say, Baytu 1, 2, Kupu- 
khta, Shandasha 1, 4, Tursumbay, etc. (Kuz'mina 1962a; 1963b; 1964a; 1964b; 1965a, 
etc.). Surveys were also conducted in western Kazakhstan, the Orenburg, Chelyabinsk and 
Kurgan regions in the Urals where we discovered or made secondary investigations of 
about 100 sites, and excavated cemeteries at Emba, Kozhumberdy, Tuktubaevo, Kin- 
zerskiy, and Alakul', and the settlement of Chernorech'e (Kuz'mina 1961b; 1969; 1973a; 
etc). From 1961 we began conducting research expeditions in Central Asia, in Tadzhi- 
kistan headed by M. M. D'akonov and A. M. Mandel'shtam; in Uzbekistan headed by Yu. 
G. Gulyamov, in Turkmenistan headed by A. M. Mandel'shtam, A. A. Marushenko and 
V. I. Sarianidi. The author also became acquainted with the excavations conducted by G. 
B. Zdanovich in the Urals, G. B. Zdanovich and M. K. Kadyrbaev in Kazakhstan, P. N. 
Kozhemyako in Kirgizia, N. G Gorbunova in Fergana, B. A. Litvinsky in Tadzhikistan, 
M. A. Itina in Khorezm, etc. 



CLASSIFICATION OF SITES AND PRIMARY FEATURES 1 9 

The relative chronology of each type was defined on the basis of: 1) strati- 
graphic evidence of cemeteries and settlements; 2) series exhibiting continuity of 
typological development; 3) type co-occurrence; and 4) synchronization with 
stages of other cultures according to imports in closed complexes and assessment 
of the absolute age of every site within the single long chronological system 
developed for the steppes by S. V. Kiselev (1960), N. Ya. Merpert (1961a), A. I. 
Terenozhkin (1965) and V. S. Bochkarev and co-ordinated with the European 
schemes of H. Muller-Karpe (1959; 1960; 1980) and W. A. von Brunn (1959), 
with the long chronology of China, and with that of south Central Asia according 
to V. M. Masson (1956; 1959) which is based on Hissar 3. The radiocarbon se- 
quence for Andronovo is a matter of heated debate due to the wide variability in 
the range of dates (see Appendix I). 

I completed my classification of the sites in 1981 (Kuz'mina 1982; 1986a, b; 
1994) which was published in the monograph (in Russian) "Earliest Cattle- 
Breeders from the Urals to the Tian-Shan" (Kuz'mina 1986b). Great attention 
was paid in it to the analysis of metal objects that accompanied cheek-pieces (see 
chapter 8) as support for the chronological scheme. 

Since 1981 many new sites have been excavated and these have provided an 
opportunity to verify the proposed scheme. The major site types presented in the 
analysis comprise the Sintashta, Petrovka, Alakul', Fedorovo, and a large 
number of Timber-grave and Andronovo sites which were found around 
Magnitogorsk, the southern Urals and in Central Asia. 



Petrovka 

Sites of the Petrovka type are found in the southern Urals, and northern and 
central Kazakhstan (Map 5, Figs. 14, 78). These comprise the following 
cemeteries: Novy Kumak, Ibragimovo, Stepnoe I, Troitsk, Tsarev Kurgan in 
Kurgan, Chaglinka, Grafskie Razvaliny, Raskatikha, Nikolaevka II, Petrovka, 
Bakteniz, Aksayman, Kenes, Novonikol'skoe, Ulyubay. The Petrovka sites are 
Nurtai, Satan, Ak-Tobe I, II, Krasnaya Krucha, Zvenigorodka, and the 
settlements of Semiozernoe, Kulevchi III, Zhelkuar, Konezavod III, Petrovka II, 
Bogolyubovo, Novonikol'skoe I, Amangel'dy I, Kenotkel V, Ikpen' I (Lento v- 
sky 1929; Formozov 1951a, 120-121; Semenov 1956a; Matveeva 1962; Stokolos 
19662b; Sal'nikov 1962a, 41; 1967, 33; Potemkina 1969; 1983a; 1983b; 1985 
Zdanovich 1973a; 1973c; 1974b; 1975; 1976a; 1983; 1988, 22-57, 71-86; 1989 
1992; Zdanovich G and S. Zdanovich 1980; Smirnov and Kuz'mina 1977 
Evdokimov 1980a; 1984; Vinogradov 1982; 1983; Ivanov and Ismagilov 1981, 
133; Tkachev 1987; 1991; 1993; Kalieva et al. 1992). 

The cemeteries are of three types: children's graves, flat-grave cemeteries 
(Raskatikha, Petrovka) and cemeteries of a small number of rather low earthen 
kurgans with one or two central pits and sometimes with adult and children 
burials in individual small pits in a circle (types 1.1 and 1.2 [Fig. 1]). Graves are 
large, sometimes with a step and wooden constructions, often covered with clay, 
and traces of fire; in central Kazakhstan there are burials in stone cists. The 
deceased are positioned on the back with knees raised or on the side slightly 
flexed; orientation is variable, more often westwards and north-westwards but 



20 CHAPTER THREE 

also eastwards and north-eastwards. There are animal bones in the grave, on the 
covering or in a sacrificial complex. There is also a group of graves with chariot 
burials, harnessed horses (sometimes bridled), a rich set of weapons, and the 
skulls and feet of cattle and sheep (Figs. 35:1; 68). 

Stratigraphically, the Petrovka complexes follow the Sintashta ones, which 
are based on the Catacomb (Novy Kumak) culture(s), and these are followed by 
the Alakul' complex (Tsarev Kurgan, Novy Kumak Stepnoe I, Kenes, Grafskie 
Razvaliny, Novonikol'skoe, Petrovka II, Kulevchi III) or Fedorovo (Novonikol'- 
skoe, Ikpen' I) burials. They date to the 16th century BC according to the finds 
of cheek- pieces that are analogous to those recovered at Mycenae (Figs. 36, 37). 
Archaic cheek-pieces and forms of pre-Seyma metal types (Figs. 29, 30, 33, 39) 
plead for a date in the 17th century BC. 

Petrovka knives/daggers, adzes, and open-socketed spears derive from 
Sintashta and Abashevo and they typologically precede Alakul'. Similar to 
Abashevo are the single-bladed knives, bone rod with plug (Fig. 39: 23) 
pendants with one and a half spirals, bracelets with open ends, glass-like 
pendants, and headgear consisting of chains of beads, single beads, and 
perforated pendants and plates. Bronze warty beads from Sintashta, Alabuga and 
Grafskie Razvaliny have correspondences in the northern Caucasus and the late 
Catacomb Don area (Bratchenko 1976). Axes with a butt in the shape of a firing 
pin from Sintashta and Berezovka are similar to those from Malinovka that are 
assigned to the Zajmishche stage of the Pre-Kazan' culture of the 16th century 
BC according to A. Kh. Khalikov (1969). Stone arrowheads with truncated base 
derive from the Catacomb (Bratchenko 1989) culture(s) and are known in the 
Abashevo and early Pre-Kazan' complexes. A stone axe from Krasnaya Krucha 
(Fig. 39: 21) is archaic in form and close to those found in Fatyanovo (Krainov 
1972: 57) and Catacomb (Bratchenko 1976: fig. 72, 16), a burial in Khashchevoe 
(Kovaleva et al. 1979; fig. 3, 2), and Abashevo (Podkletinskiy burial, Shilovskoe 
settlement) (Pryakhin 1976: fig. 9, 25; 1977: fig. 8,1) and partially early Timber- 
grave (Podstepki, Kordon Derkul'skiy) (Agapov 1977, fig. 1; Rykov 1927b: fig. 
17) cultures, which establishes its date as not later than the middle of the 2nd 
millennium BC. Stone maces and arrows have a wide range of occurrence, but 
do not contradict the dating suggested here. 

The Petrovka site type includes related Potapovka complexes of Bashkiria 
(Almukhametovo), the Volga area (Alekseevka II, Potapovka, Utevka VI) and 
the Don area (Vlasovskiy, Pichaevskiy and Kondrashkinskiy) (AO 1911: 150, 
198; 1980: 133; Potemkina 1983a; Vinnikov and Sinuk 1990; Moiseev 1990; 
Pryakhin 1992; Vasil'eve? al. 1992; 1994; 1995). 



Alakul' 

The Alakul' type is situated in the Chelyabinsk and Tobol'sk regions (the 
cemeteries of Alakul', Chernyaki I, II, Isakovo, Kurgan 15, Tsarev Kurgan, 
Baklanskoe, Churilovo, Kamyshnoe, Verkhnyaya Alabuga, Subbotino, 
Alekseevka, Pereleski, Evgen'evka, Chistolebyazh'e, Khripunovskiy), in Uysko- 
Uvel'ka (Chernoozer'e I, Berezovskiy, Stepnaya Kommuna, Biryukovo), in 
Magnitogorsk (Tamerlane Tower), in north Kazakhstan, in the steppe zone 



CLASSIFICATION OF SITES AND PRIMARY FEATURES 2 1 

(burial at Kokchetav, Efimovka, Borovoe, kurgan 1, Nurmambet) and in the 
forest steppe (Semipalatnoe). We distinguish pure Alakul' burials in western 
Kazakhstan from the Kozhumberdy ones (burials in Orsk Nikel, Ulke I, 
Aktyubinsk Poligon and Ptitsefabrika, Emba), in central Kazakhstan from the 
Atasu type which were formerly viewed by Kazakhstan archaeologists as a local 
variant of Alakul' (Batkin Paek Egiz-Koytas, Bylkyldak III, Karasay, Begazy, 
Ograda I, Karabie, Tash-bulak, Algabas, at Kirov Sovkhoz stone circle). A. A. 
Tkachev (1987; 1991) assigned the cemeteries in Sary-Arka to the Alakul' type, 
i.e., Kopa, Kopa I, Achshi-Ozek Nurken, Tashik Shapat, Maitan and to the late 
Alakul' were assigned Bozingen and Izhevskiy I. It is difficult to draw a firm 
conclusion before complete publication, but the existence of Fedorovo and 
mixed types of vessels from the sites, as well as the presence of annexes to the 
enclosing fence of the kurgans and instances of cremation suggest that at least 
part of these sites belong to the Atasu type according to my periodization. The 
most eastern Alakul' site is the Ermak cemetery near Omsk. 

Alakul' sites contain up to several dozen constructions: in the forest -steppe 
zone we find kurgans with subsoil burial chambers; in the steppe zone the 
kurgans are surrounded with a stone ring (cromlech) constructed from vertically 
set stone slabs (types I l5 I 2 ; Ha, Ilia). In the center of the construction there is a 
large burial pit; in the forest-steppe we sometimes encounter children burials as 
secondary additions surrounding the main burial or inserted into it (I 2 ); in the 
steppe there is sometimes a child grave within a stone circle. A cemetery of 
children buried in subsoil graves is known from Baklanskoe. 

The graves are rectangular in shape. The material of construction varies: in 
the forest-steppe they are subsoil chambers, often constructed with a timber 
framework or faced with split timber logs; in the steppe stone cists are more 
often found. The tombs are roofed with layers of wood or planking which are 
absent from the steppe burials where roofing was effected with stone slabs. In 
general, inhumation burials are flexed on the left side and only rarely on the 
right. The orientation was generally to the west, more rarely to the south-west; 
exceptions to the general rule are noted: at Chelyabinsk the burials are oriented 
to the south or follow no obvious rule; in Magnitogorsk orientation is to the 
north; in Tobol'sk and Sol'-Iletsk it is eastern, a reflection of the influence of the 
Timber-grave rite. Paired male and female burials are known, more rarely are 
paired burials unisex or involve mother and child. At the head of the deceased 
there are two vessels, rarely more, and sometimes a single vessel in a child's 
grave. 

Women are often buried in ceremonial attire: a dress sewn with plates, shells, 
claws, bronze and paste beads, with bracelets and finger -rings, beads sewn on 
boot-straps, temple ornaments or earrings, and decorations for braided hair (Fig. 
33). A bronze knife and weapon (socketed spear, stone, bone and bronze arrows, 
stone or bronze axe, mace or adze) accompanied the male burials (Fig. 39, 40). 

A characteristic ritual feature was the sacrificial placement of a horse skull 
and legs (head-and-hooves), more rarely an ox or ram, on the roof of the grave or 
in the kurgan mound. Burials of dogs are known (Alakul', Chistolebyazh'e, 
Chernyaki). In Alakul', Khripunovskiy and Maitan there are pairs of chariot 
harnesses; in Maitan there are also cheek-pieces. Some vessels and animal bones 
constitute a funeral feast. 



22 CHAPTER THREE 

Marked differences between Alakul' sites are conditioned by: 1) their partic- 
ular environmental zone which influenced domestic architecture and then the 
burial constructions that imitate them (e.g., the presence of stone constructions in 
the steppe); 2. substrates and neighboring cultures whose influence can be seen 
in deviations from standard grave orientation and ceramics (tempering agents, 
form, methods of ornamentation). A combination of these features account for 
the peripheral regional variants, i.e., Chelyabinsk, Tobol'sk, Uysko-Uvel'ka, 
Magnitogorsk, north, west and central Kazakhstan (Fig. 4). 

Monuments of the Alakul' type derive from the Petrovka, sharing with them 
basic defining cultural features — burial rite, technology, ceramic form (Fig. 4) — 
that prove their direct genetic succession (Fig. 3) as well as ECT features: 
economy, architecture, metal objects, cheek-pieces, and finally, ornament. There 
is a large group of transitional complexes: Pyatimary, Gerasimovka I, 
Berezovskiy, Stepnoe, Kulevchi VI, Raskatikha, Alabuga, Chistolebyazh'e, 
Kenes, Ulke I, Orsk, Krasnaya Krucha, Nurtai with transition-type vessels, 
including vases with a rib under the rim, archaic types of knives (Ulke I, Nikel, 
Krasnaya Krucha (Fig. 30)) and axes of bronze (Berezovskiy) and stone 
(Krasnaya Krucha, Kokchetav, Pyatimary) (Fig. 39). 

The chronology of the Alakul' type is established on the basis of: 

1. The stratigraphy of a) burials: in Tsarev Kurgan, Grafskie Razvaliny and 
Kenes where the main graves are of the Petrovka type and are cut by Alakul'; b) 
settlements: at Novonikol'skoe, Petrovka II and Kulevchi III, Alakul' construc- 
tions cut the Petrovka layer, and the first two sites were then covered by a Final 
Bronze layer. 

2. Synchronization between the Timber -grave and Alakul' cultures. This is 
based on burials of the Timber-grave and Alakul' in the contact zone (Novobelo- 
gorka, Gerasimovka I, II, Burdygino, Spasskoe I, II, Agapovka II, Maly Kizil I, 
II,); and the discovery of Alakul' vessels and their imitations in Timber-grave 
burials of Stage 2 (Novo-Baskakovo, Komsomol'skiy III and IV, Bikeshevskiy 
IV, Pestrovskiy, Tavlykaevo III and IV in Bashkiria, Staroivanovka, Zolotaya 
Niva in the Samara Volga (Morozov and Pshenichnyk 1976, figs. 3: 2, 3; 4: 1-3, 
9; 5: 2, 6-10; Zudina 1981, figs. 2: 2, 4; 3: 2, 4; 6: 5, 6; Kuz'mina 1987b; Rutto 
1989). 

3. Securely dated metal artifacts recovered from closed complexes of Alakul' 
burials. These include wide axes with a broad butt (Fig. 40: 1) (Tsarev Kurgan, 
Evgen'evka, Bozingen), adzes (Orsk, Emba), spears (Orsk), double-edged 
socketed knives (Tsarev Kurgan, Evgen'evka, Alabuga, Khripunovskiy, 
Chistolebyazh'e, Pereleski, Subbotino, Orsk, Borovoe, Semipalatnoe, Maitan, 
Bozingen), knives with soldered handle (Chernyaki, Nurmambet) assigned to the 
Seyma chronological horizon and that are both typologically and technologically 
(mainly cast in bivalve forms) derived from Petrovka and prototypes of Final 
Bronze Ages forms (Figs. 29, 30, 32, 39, 40). They date to the 15th-14th 
centuries BC on the basis of analogies with East European cultures. Alakul' plate 
and disk shaped cheek-pieces without tenons from Alakul' and Novonikol'skoe 
are dated to the 15-14th centuries BC, and are comparable to cheek-pieces from 
the Danube of the Toszeg and Vatina type (Figs. 37, 39). 



CLASSIFICATION OF SITES AND PRIMARY FEATURES 23 

Fedorovo 

The Fedorovo type is represented in the Urals (Fedorovo, Nurbakovo, Novo- 
Burino, Bol'shaya Karabolka, Smolino, Sosnovskiy, Sineglazovo, Sukhomesovo, 
Solntse-Talika, Isakovo — kurgans 2-8, Tuktubaevo, Urefty, Kinzerskiy, Urazae- 
vo), in north Kazakhstan (Borovoe, Obaly, Biyrek-Kol', Kalachevskiy, Burluk, 
Pavlovka, and probably Koshkarbay, Kenotkel', Sokolovka), in central Kazakh- 
stan (Burgulyuk I, Kanatas, Sangru II, Akshatau, Baybala I, Kosagal, Botakara, 
Dzhamantas, Myrzhik, Alypkash, Dandybay 2-8, some of the burials at Zhilan- 
dy), in east Kazakhstan (Maly Koytas, Marinka, Karausek, Sarykol I, II, Zhana- 
zhurt, Barashki, Berezovskiy, Kyzyltas, Betkuduk Belokamenka, Dzhartas, Pred- 
gornoe, Semipalatinskie Dyuny and probably some of the graves found at Kanay, 
Karadzhal, Oblaketka, Kyzyltau), in Pavlodar near the Irtysh (Akmola, Lebya- 
zh'e), on the Upper Ob (Zmeevka or Krasny Yar, Khomutinka, Shipunova, Ikon- 
nikova, Blizhnie Elbany 12 and 14, Ordynskoe, Novoaleksandrovka, Nizhnyaya 
Suetka, Elunino, Gryaznovo, Kytmanovo, Ur, Elovka 2, Vakhrushevo, Mikhay- 
lovka, Preobrazhenka 3, Sopka 2, Bol'shepichugina, etc.), on the Upper Yenisey 
(Andronovo, Orak, Uzhur, Pristan, Solenoozernaya, Sukhoe Ozero, Novaya Che- 
rnaya II, III, Yarki I, II, Ust'-Erba, Kamenka II, III, Lanin Log, Podkuninskiy, 
Ashpyl, Kadat, etc.), in the Tian-Shan (Arpa, Prigorodnoe, Issyk-Kul'), in the 
Pamirs (Kyzyl-Rabat, Kokuybel'su, Yuzhbok) and in the interfluvial region of 
Central Asia (burial at Tandyryul, settlement and collections at Kirov Sovkhoz 
(Fig. 83), Dzhilikul', Karabura, and Saksanohur (Maps 2-4, Fig. 19-21, 41). 

Cemeteries contain from ten to several dozen burial structures (sometimes as 
many as 150). These include kurgans, often enclosed by a circular or square 
fence; chambers built of dry-stone walling or, more rarely, as cists from upright 
slabs; fences (types I, Ha, lib, Ilia, Illb, IV) and also multiple chambers covered 
with kurgans and enclosed by fences (types V-VIII); there are also flat 
cemeteries (Ob). Children burials are known (Yarki, Sukhoe Ozero la), 
including both flat graves (Elbany, Zmeevka) and special kurgans (Novaya 
Chernaya III, 24). 

In the center of the construction is one large square or rectangular flat grave 
constructed as a cist, timber -construction, block facing, within a stone fenced 
enclosure, rarely with a stone cist. 

The burial rite comprises cremation (where the burning was not carried out in 
situ) and inhumation. Cremation dominates in the Urals; in central and northern 
Kazakhstan the cemeteries are bi-ritual; in eastern Kazakhstan and south Siberia 
inhumation prevails; children are mainly inhumed. The general Andronovo 
burial orientation is western, sometimes south-western. Paired male and female 
burials and mother and child burials are known in the east. There are two vessels 
in the grave (rarely more; in children graves there is sometimes only one vessel); 
in the Urals a dish is often found. Ornaments comprise temple rings with a 
narrow return or funnel-shaped opening and single round plates (Fig. 33); other 
goods are extremely rare except in some late cemeteries on the periphery of the 
Andronovo culture. 

Ritual features attending the burials includes placing ribs, more rarely the 
shoulder-blade, of a horse (sometimes the shoulder-blade is of an ox or ram); 



24 CHAPTER THREE 

head-and-hooves combinations are rare. Dog burials are also known (Fedorovo, 
Sukhomesovo, Kinzerskiy, Sukhoe Ozero, Ust'-Erba). 

The Fedorovo type is divided into a series of local variants according to 
certain specific features. These variants comprise the Urals, central, northern, 
eastern Kazakhstan, the Ob, the Yenisey, and Central Asian (Fig. 6). The further 
one moves from central Kazakhstan the frequency of the complex diminishes 
and substratum elements increase; they are especially marked in the Yenisey 
region (Ust'-Erba, Sukhoe Ozero, Novaya Chernaya) where Okunevo cultural 
traditions can be seen in fences, constructions surrounding graves, square vessel 
forms and ornaments. 

K. V. Sal'nikov saw the sites of the Fedorovo type as an early stage of the 
Andronovo culture while N. A. Avanesova and G. B. Zdanovich placed them in 
a late stage. We have been able to classify them into two temporal groups (Fig. 

5). 

There is no stratigraphic correlation between the Fedorovo and Alakul' types 
excepting the possibility of some stratigraphy from the settlement at Kipel'. At 
the settlement site of Novonikol'skoe the Fedorovo layer lies over the Petrovka 
and under the Alekseevka, and according to T. S. Malutina (1991) there is a 
synchronous mixing of Fedorovo-Alakul' ; at Atasu, Ust'-Kenetay, Rodionovka 
and others, Fedorovo is found under Alekseevka; in Korchazhka V Fedorovo is 
under Irmen; in Ikpen' (Fig. 80), the lower part of an ash-pit contained late 
Petrovka vessels, the middle of the pit held Fedorovo remains, and the upper 
Alekseevka (Tkachev 1991). There are also unstratified findings of Fedorovo 
vessels on Petrovka type sites such as Petrovka, Kulevchi and Sintashta, that hint 
at an early dating of the Fedorovo sites. At Sintashta there are vessels that may 
be seen as proto-Fedorovo with their vase-like profiles, oblique triangle 
ornament, and hanging festoons. An Afanas'evo sherd was found in the fill of a 
Fedorovo grave at Ust'-Erba; there is a Fedorovo pot in a Karasuk grave. In 
Sukhoe Ozero a Fedorovo cist was covered by a secondarily used Okunevo stela; 
at the settlement of Itkul the Fedorovo layer was covered by Samus' material; in 
Preobrazhenka III Fedorovo graves cut the Krotovo layer and were covered by 
kurgans of the Irmen group. In Siberia, therefore, the Fedorovo type follows the 
Afanas'evo, Okunevo, Krotovo, and Samus' cultures, and is covered by the 
Karasuk-Irmen complexes. 

Early Fedorovo sites date to the 15th 14th centuries BC according to the 
evidence of: 1) the technology and typology of a few metal objects — grooved 
knives, analogous to types found in the Alakul' and Timber-grave groups 
(Urefty, Koytas, Semipalatinskie Dyuny), flat spiral bracelets, with analogs in 
Alakul' (Subbotino), forged temple rings and plates (Fig. 33); 2) Fedorovo 
ceramics that have been found associated with celts (Fig. 31), knives, spears with 
sleeves that were produced by Fedorovo metallurgists of eastern and central 
Kazakhstan and found in pre-Andronovo sites of the west Siberian Samus' and 
Krotovo cultures (Samus', Preobrazhenka III, Sopka II, Elunino; Fig. 41); 3) 
finds at Karluga of a three-hole rodlike cheek-piece analogous to those found in 
the Toszeg and Ottomani cultures (Milller-Karpe 1980: tables 290: A26, 291: 
A21); 4) the discovery at Korkino of a unique stone shaft-hole axe with a blade 
combining features of axes I and IV from the Borodino hoard (Fig. 39); 5) the 
participation of Fedorovo tribes in the formation of the Tazabagyab and Vakhsh 



CLASSIFICATION OF SITES AND PRIMARY FEATURES 25 

cultures of Central Asia; 6) Fedorovo cremation graves found under the fill of 
contemporaneous Alakul' and Timber-grave- Alakul' kurgans and the presence 
of Fedorovo vessels in closed complexes of Alakul' graves and vice versa 
(Chernyaki II, l/3,4;4; Subbotino 2/4, 3; Urefty, Nurbakovo 2; 3; Isakovo 1, 2, 4, 
5; Myrzhik Semipalatnoe, Sukhomesovo 3; Sosnovskiy 1; Kinzerskiy 1, 32; 
Priplodny Log, Burluk, Alypkash, Spasskoe I, l;2/5; 4; Spasskoe 11,1; Agapovka 
2/5; 4/1; 7). The discovery of Fedorovo vessels and their imitations in graves of 
stage II of the Timber-grave culture (15th— 14th centuries BC) of the south Urals, 
Volga, Don, and Ukraine regions; and 8) the presence of Fedorovo pottery on the 
Shortughai settlement in Afghanistan in the post-Harappan layer. 

For late Fedorovo cemeteries we have Tuktubaevo, Smolino, Urazaevo II, the 
periphery of Kinzerskiy, Priplodny Log, Biyrek-Kol', Sangru II, Barashki, Bere- 
zovskiy, Zevakino, Predgornoe, Elovka II, Blizhnie Elbany, Suetka, Kytmanovo, 
and the cemeteries on the Yenisey. Typical here are elongated kurgans and 
enclosing fences (types V, VI, VII, VIII) containing several successive burials; 
sometimes in the east the graves are narrow and shallow, sometimes burial 
chambers erected at ground level are in the form of cists with buttresses (Fig. 1); 
the vessels are of type II with disrupted and changed proportions and bulging and 
swollen body, the zonal decoration has been replaced and is absent from the 
lower part of the vessel and is sometimes executed in two zones (Figs. 5, 11); 
pots of the second type have a wide base, narrow neck, are poorly decorated or 
undecorated; sometimes in the peripheral areas there is a rich selection of grave 
goods. 

Late Fedorovo sites date to the 13th century BC and later. This can be argued 
on the basis of: 1) their stratigraphic position below the Final Bronze Age layer 
on settlements; 2) Fedorovo pottery in a Karasuk grave (Ust'-Erba); they corre- 
late with a later intrusive grave or added fences: Karasuk (Orak probably Kyt- 
manovo), Elovka-Irmen (Elovka, Elunino), and the Early Iron Age, with knives 
of 8th-7th centuries BC (Zevakino) that serve as a terminus ante quern for 
Fedorovo; 3) metal objects that represent a Fedorovo development in technology 
(high-tin bronzes, cast in bi- and tri-valve molds), and types with analogies in the 
Karasuk culture and Final Bronze Age hoards: knives with a rhomboid cross- 
section, with a ridge or linear rib (Zevakino, Elovka), single-bladed knives 
(Elovka; Fig. 30), arrows with a short socket (Smolino, Biyrek-Kol', Elovka), 
plates with a loop (Elovka, Orak, Sukhoe Ozero), mirrors with a loop (Sukhoe 
Ozero, Buguly, Elbany), notched spears and flanged sickles (Predgornoe), brace- 
lets with spiral cones (Sangru, Aleksandrovka, Kytmanovo, Elovka, Suetka), 
octogonal plates (Elovka), cast temple rings (Elovka, Orak), etc.; 4) associations 
with Cherkaskul' vessels (Tuktubaevo, Priplodny Log); 5) the influence of the 
Begazy culture on the construction of burials (Priplodny Log); 6) presence of 
Fedorovo material in a layer of a late Timber-grave settlement of stage IV on the 
Volga; and 7) synchronization with farming complexes of the late (Mollali) stage 
of the Namazga VI culture; late Fedorovo and wheel-made ceramics were found 
together in north Kazakhstan on the settlement of Pavlovka and in Central Asia 
in cemeteries at Tandyryul, Kumsay and Dashti-Kozi and on settlements at 
Kangurt-Tut and Teguzak (Fig. 48). 

Classification of the material has led to the definition of two clear site 
types — Alakul' and Fedorovo — and it has established their contemporaneity over 



26 CHAPTER THREE 

the main part of their distribution. The main distinctions are in burial 
constructions (kurgans with stone kerb versus cisted enclosures without a 
kurgan), graves (timber-built chamber versus stone cist); burial (inhumation 
versus cremation); animal sacrifice (horse burials or their skins with skull and 
legs versus only ribs) and, what is especially important, the technology of 
production, form and principles of ceramic ornamentation (ceramics built from 
the outside versus those built from the inside; ribbed shoulder versus rounded 
shoulder; ornament applied on a square lattice versus an oblique lattice; even and 
large toothed stamp versus small toothed stamp (Fig. 2.11-13). We thus we do 
not see the Alakul' and Fedorovo types as two genetically connected stages of a 
single culture. The ethnographic importance of two different ceramic traditions 
and burial rites leads us to see them as two independent lines of development 
that reflect two genetically different population groups. However, we cannot 
agree with V. N. Chernetsov and his followers who view them as ethnically 
distinct, one Iranian and the other Ugrian, but rather as closely related groups of 
a single ethnic unity; this is demonstrated by the integrity of the majority of 
ethnically meaningful features: the similarity of many of the types of burial 
constructions, western orientation of the graves, tripartite structure of pottery, 
three-zone decoration, common ornamental elements, etc. 

Numerous complexes are distinguished over a large part of Andronovo 
territory in which the Fedorovo and Alakul' traditions are found in different 
combinations. We view them as the product of integration and assimilation, 
occuring at different times and in different territories and leading to different 
results; this view contrasts with that of K. V. Sal'nikov, N. A. Avanesova and G. 
B. Zdanovich who cite mixed monuments of the Kozhumberdy type as a proof of 
a transitional phase from one stage to another. Syncretic sites are distinguished 
by marked local peculiarities that permit their typological classification. 



Sol'-Iletsk 

The Sol'-Iletsk type of the Alakul' line of development is localized in the 
southern Urals (the cemeteries of Uvak, Mechet-say, Pyatimary, Bliznetsy, 
Krasnopartizanskiy, Vetlyanka, Vetlyanka IV, Dolgoe, Peshchanoe Ozero; Map 
2, Fig. 24). Here we find subsoil kurgans with large central graves, covered by 
stone slabs or layers of wood; inhumation is with the head to the west, rarely 
south-west. In Mechet-say and Vetlyanka IV there are also two instances of 
cremation. Vetlyanka is a subsoil burial with a large grave in the center and num- 
erous graves on the periphery; this along with its north-east orientation reflects 
Timber-grave influence. In Uvak in kurgan 15 there are 19 burials, mainly 
children, under one mound. Now and then there are sheep and cow bones and at 
Vetlyanka IV there are the remains of a horse with a chariot (?) (Gorbunov, 
Denisov and Ismagilov 1990). The fabric of the ceramics reveals an admixture of 
shell and sand; incised surfaces reflect Timber -grave influence. The pots are 
often with a narrow neck (variant B), with or without a ribbed shoulder; orna- 
ment is over the rim, neck and shoulder (absence of decoration on the neck is 
rare); the ornament is arranged over a square lattice; a specific ornament is a 
band with connecting rhombuses (Fig. 24). The type emerged very early which is 



CLASSIFICATION OF SITES AND PRIMARY FEATURES 27 

indicated by vessels with hanging rim and graves with a bench; it is dated to the 
15th 14th century BC according to a stone axe of the Borodino type (Pyati- 
mary), an archaic knife and Seyma-type spear (Bliznetsy), a knife, flint arrows, 
stone and bronze ones with rolled up socket, bone bead (Uvak), a knife, flint and 
bone arrows (Vetlyanka IV; Figs. 30, 32, 39). 



Kozhumberdy 

The Kozhumberdy type of the Alakul' line of development is found in western 
Kazakhstan (the cemeteries of Tursumbay I, II, Elenovka II, Shandasha I, II, Ku- 
pukhta, Baytu I, II, Ataken-say, Ushkatta, Tasty-Butak, Buget II, Kozhumberdy, 
Kirgil'da I, II, Tulaykin Aul, Novy Akkerman, Khabarnoe I, II, Kunakbay-say, 
Ural-say, Pochtovy Post, Baturasay, Ilek, Rossovkhoz, Kuagash, and the settle- 
ments of Dzharly and Elenovka microdistricts, Tasty-Butak). The kurgans are 
characterized by having a stone circle (types Ha, Ilia) and in later kurgans by 
adjoining fenced enclosures (type Vila). Graves are subsoil and in cists. Burial is 
by inhumation with head to the west, rarely south-west; charcoal is present. 
There are combinations of animal skulls and ribs. Ceramic technology is 
syncretic with textile patterns and applied bands; pots both with and without a 
ribbed shoulder; ornament may be over a square or oblique lattice with necks 
void of decoration; specific ornamental technology involves oblique ornamental 
bands and the use of large-, medium- and small-toothed stamping (Fig. 23a, b, 
24). The type evolves over a long time and can be divided into an early Ushkatta 
and later Shandasha stage (Kuz'mina 1963d; 1965a). The type is dated from the 
16th century BC, and its main floruit is from the 15th through 13th centuries BC. 
This can be demonstrated by the knives (Kozhumberdy, Kumak, Kupukhta, 
Bayturasay, settlements of Ushkatta and Shandasha), by arrowheads (Tasty- 
Butak), by cheek-pieces (Tasty-Butak; Figs. 30, 39) and by ceramics and 
ornaments from Timber-grave complexes of the II and III stages in the Volga 
area (Kuz'mina 1987b) and from the Tazabagyab settlements of Kokcha 15, 16, 
Dzhanbas 34 (Itina 1977, figs. 24, 39, 40, 57, 59). 



Atasu 

The Atasu type is found in central Kazakhstan (cemeteries: Atasu, Ayshrak, Ak- 
Mustafa, Koyshoku, Bylkyldak I, II, Shet, Murza-Shoku, El'shibek, Aksu- 
Ayuly, Zhilandy, Zhaman-Uzen II, Lisakovskiy, Bes-Oba, Altyn-Tyube, Nurken, 
in the Irtysh area Balakty). A. A. Tkachev (1991) assigned the cemeteries of 
Bozingen and Izhevskiy I to the late Alakul' stage; E. F. Usmanova (1987) 
assigned the Ayshrak and Belakty burials to the Alakul' type as there were no 
clear Fedorovo features in these complexes. Large cemeteries are characteristic 
of the Atasu type. They contain up to several dozen constructions (Ayshrak — 
100) and are enclosed by round, oval and rectangular fences (IVa, Vila, Villa), 
often with two or more stone cists in the center. Burial is by inhumation, head to 
the west or south-west; there are several instances of cremation. There is a 
combination of two ceramic traditions: vessels in the Alakul' style built from the 



28 CHAPTER THREE 

base upwards with bands of clay on the outside and the Fedorovo method, with 
the bands built up from the interior and a footed base. Vessel types are 
numerous: with or without a ribbed shoulder, with wide or narrow neck (variants 
A and B), often with a high rim and sometimes with a footed base. Ornament is 
on the rim, neck and shoulder; it is arranged over a square lattice or, more often, 
an oblique lattice. There are various ornamental motifs and combinations; 
ornamental bands are horizontal. This type dates to the 15th— 13th centuries BC 
according to: 1) the stratigraphy of the settlement at Kopa, where the Atasu layer 
is covered by a Final Bronze Age one; 2) the technology and typology of the 
knives (Bylkyldak Altyn-Tyube (Fig. 30), arrows (Ayshrak Zhaman-Uzen II, 
Koyshoku, Shet, Karasay II), bracelets with protruding spirals (Ayshrak), that 
directly preceded Final Bronze Age ones; paw-shaped pendants (Aksu-Ayuly, 
Koyshoku) analogous to those found in the Karasuk culture; ornaments of round 
plates (Fig. 39) imitating Bactria and Margiana seals. A spade-like pin from 
Ayshrak ties the Atasu type with the farming cultures of south Central Asia. 

The Fedorovo complexes of central Kazakhstan are purely Fedorovo while 
individual Fedorovo elements are known from late Petrovka and Alakul' sites. 
The Bronze Age complexes of central Kazakhstan are extremely numerous and 
reflect different combinations of Fedorovo and Alakul' components (which 
sometimes makes their classification difficult), unlike the more homogenous 
Sol'-Iletsk and Kozhumberdy sites. This probably indicates that the assimilation 
process in the southern Urals and in western Kazakhstan finished early (by the 
end of 16th century BC on the basis of Sintashta type sites?) and a newly formed 
population evolved in the 15th— 13th centuries BC. In central Kazakhstan the 
active interaction of different groups of populations took place for a long time, 
leading to various forms. This historical situation was probably preconditioned 
by the rich copper resources of the region, especially tin, which attracted many 
tribes there. In any case, new materials in central Kazakhstan, previously the far 
eastern border of the interaction of Fedorovo and Alakul' tribes, provides 
conclusive support for the different geneses of these two groups of populations. 



Amangel'dy 

The Amangel'dy type of the Alakul' line of development is situated in north 
Kazakhstan (cemeteries: Amangel'dy, Petropavlovsk, Aydabul, Kuropatkino). 
Burial was in earthen kurgans, sometimes with an enclosure and adjoinging 
fences (types la, Ha, Va); there are also timber-constructions. Inhumation burial 
was with the head to the west. Burials were accompanied by the skulls and legs 
of animals. Pots had ribbed or rounded shoulders; the neck was often without 
ornament; the ornament was often over an oblique lattice, rarely over a square 
lattice; vessels were with or without shoulder ridges; ornament: bands are 
horizontal and were made by continuous impressions from a smooth comb and 
medium- and small-toothed stamping; specific ornament includes high triangles 
and zigzags on the neck. The type dates to the 15th 13th centuries BC on the 
basis of a cheek-piece (Aydabul; Fig. 39) and arrows (Amangel'dy). 



CLASSIFICATION OF SITES AND PRIMARY FEATURES 29 

Tautary 

The Tautary type belongs to the Fedorovo line of development and is localized in 
south Kazakhstan (Tautary, Kuyukty). Burials are surrounded by square, 
rectangular, and sometimes round enclosures (types VI, VIII, III). Burial is in 
subsoil graves, usually cremation, rarely inhumation. There is a combination of 
two ceramic traditions (Fig. 27): pots may be with or without a ribbed shoulder; 
the ornament is applied against a square or oblique lattice; specific traits include 
negative filling of the ornament and degraded ornament. The type dates to the 
12th century BC and later according to the presence of a cist lying on a clay 
foundation (Kuyukty), characteristic of the pre-Begazy architecture, a Karasuk 
paw-shaped pendant (Tautary), and vessels imitating pottery of the late Namazga 
VI stage. 



Semirech 'e 

The Semirech'e type is syncretic and is found in Semirech'e and Fergana (Map 
4, Figs. 26; 73a, b). The type emerged out of a crossing of Fedorovo (Arpa, 
Prigorodnoe, Issyk-Kul') and Alakul' (the site at the Belovodsk citadel). To the 
early stages belong the Kapal cemetery where Alakul' features prevail in 
ceramics while Mynchunkur III and IV, Talapty I and II, Kuygan II and Kara- 
Kuduk display Fedorovo features (cists, bi-rituality, presence of vessels without 
ribbed shoulders, footed vases, and wide-necked basin-like pots). In the 
Mynshukur burial there was found a temple ring, typical for the Fedorovo type, 
with a pair of horses which have analogies in the Seyma horizon (Fig. 68). This 
helps date the type to the 15th century BC. There are no grounds for assigning 
these monuments to a special Semirech'e culture (Karabaspakova 1991: 13-14). 

A late group of Semirech'e sites are Tash-Tyube II, Tash-Bashat, Besh-Tash, 
Dzhazy-Kechu, Dzhal-Aryk, Kul'an-say, Alakul', Talapty, Kuygan, Tamgaly II, 
IV, VI, Kul'say, Uzunbulak, Kara-Kuduk, Vuadil', Arsif, Karamkul', Yapagi, 
Kashkarcha, and the settlement of Dzhal-Aryk. Graves are surrounded by fences 
with annexes (types VII, VIII) and the outlining of the pit with small stones is 
typical. The graves are subsoil, with stone boxes, small stone settings and cists 
with a clay coating. The burial ritual involves predominatly cremation but also 
inhumation. The ceramic technology is mixed: there are late Fedorovo pots with 
an expanded body and sometimes pots with ribbed shoulders; the decoration is 
not according to zones and is only found over the rim or shoulder; it is 
accomplised by a smooth, rarely large-toothed stamp; the greater part of the 
vessel is without ornament. The early type dates to the 15th— 13th centuries BC 
while the late group dates to the 12th— 9th centuries BC according to the evidence 
of the use of clay coating on the graves (analogous to Begazy); findings of a 
single-bladed knife (Tash-Tyube), arrow with a concealed socket (Vuadil') 
characteristic of the Final Bronze Age and the mutual occurence of the Chust 
culture which is dated by Iranian parallels. 

Two groups of north-east Semirech'e comprise a local variant (Karabapakova 
1987; 1989): Group I: Bigash, Aksay, Sagandy with square and trapezoidal 
boxes of the Begazy type, inhumation predominates, vessels are of Begazy form 



30 CHAPTER THREE 

with poor ornamentation; a plate with a loop was found at Sagandy; Group II: 
burials at Arsan and Bien 13. There are graves with a dromos. Pots have globular 
bodies with collared rims and poor ornament: applied rolls, nails impressions, 
nets, beads, applique knobs, characteristic of the Begazy and Dongal complexes. 
Imported wheel-made pottery is found on the Bien settlement. The complex 
dates to the 10th— 8th centuries BC and reflects the strong influence of the 
immigrant Dandybay (Karasuk by origin) population in the pre-Saka period. 



The classification of Andronovo material reveals complex historic processes of 
autochthonous development, migration and integration. It is probable that a 
migration of the Alakul' people in the second quarter of the 2nd millennium BC, 
from western into central Kazakhstan caused part of the Fedorovo migration to 
eastern Kazakhstan and south Siberia. Another part of the Fedorovo people was 
assimilated by the Alakul' people, and as a result of this integration, syncretic 
complexes were formed. Moreover, the Sol'-Iletsk and Kozhumberdy types 
formed very early: in central Kazakhstan long contacts were of different 
character, which conditioned the mixed character of the Atasu type; the 
Semirech'e type was probably formed rather late, as a result of population 
migration from central Kazakhstan. 

A review of the numerous sites of these syncretic types, united in an integral 
chain across the major part of Andronovo territory permits one, following A. A. 
Formozov, to speak about the existence of an Andronovo culture unity. 

The Andronovo materials surveyed so far provide a database for the analysis 
of the material culture of the Andronovo tribes. Of central importance in the 
development of the stockbreeding cultures of the steppe ECT was the transition 
to a nomadic type of economy, the most basic innovation of Old World culture. 
That is why the study of the ECT of this region is of primary interest for 
examining its paleo-economic development and revealing the processes 
involving the emergence of mountain pastures and then the creation of nomadic 
stockbreeding in Eurasia, which explains the dynamics and intensive 
assimilation of new territories. A. J. Toynbee, who dedicated his A Study of 
History to the philosophy of history and who greatly influenced the development 
of modern historical schools, expressed the opinion that steppe peoples were a 
catalyst of all the processes in the history of civilization. Similarly, the head of 
the French "Annales" school, F. Braudel (1969) considered the Eurasian steppe 
as a flashpoint that saw the explosion of steppe pastoralists from Germany to 
China who punctuated the slow process of cultural evolution in the Old World. 

Finally, the separate categories of the material culture of the Andronovo 
tribes and their comparison with language data and historical traditions of the 
ancient Indo-Iranians will help resolve the problem of locating their homeland, 
reject speculative constructions, and give the floor to the Aryans themselves. 



CHAPTER FOUR 
SETTLEMENTS AND DOMESTIC ARCHITECTURE 

Settlements and houses 

Numerous ethnographic and historical studies have firmly established that house 
type is conditioned not only by ecology and economy but also by the specific 
building traditions that might be preserved by certain ethnic groups over a long 
time. That is why the study of domestic architecture is of interest not only for 
economic reconstruction but also for investigating archaeological cultures and 
establishing their ethnic background. 

The discovery of the sites of the Sintashta type (Fig. 58) that belong to the 
early period of the formation of the Andronovo culture has been one of the great 
success stories of Russian archaeology (Map 12). The traditional chronology sets 
them to the 17th 16th centuries BC while radiocarbon dating places them earlier 
to the 20th-18th centuries BC (Smirnov and Kuz'mina 1976; 1977; Kuz'mina 
1984; 1998; Gening 1977; Anthony and Vinogradov 1995). 

It was I, M. Batanina (1995; Zdanovich and Batanina 1995; 1999) who, after 
examining military aerial photographs, discovered about twenty fortified settle- 
ments, the so-called "land of towns", that were situated about 40-70km from one 
other. They were found in the southern Urals in the basins of the European and 
Asian rivers, the Uy, the Ural, and the Tobol, in an area that runs 400km from 
north to south and 100-150km from east to west. These settlements are on small 
river banks, usually on promontories; they are oval, round or rectangular and are 
surrounded by defensive walls and ditches; in Ol'gino and Alandskoe the outer 
banks were strengthened by stone slabs. Associated cemeteries have been 
discovered near the settlements of Sintashta and Kuysak. The cemetery of Sol- 
ntse II belongs to the settlement of Ust'e while Bol'shoy Karagan is associated 
with Arkaim. The sites are in areas rich in oxidized copper ores, malachite and 
azurite deposits that are easy to exploit. The Vorovskaya Yama mine belongs to 
the settlement at Kuysak and Kisenet is associated with the Ust'e settlement. 

Archaeologists have excavated the settlements of Malokizil'skoe, Sintashta, 
Arkaim, Ust'e, Kuysak, Alandskoe, etc. (Gening et al. 1992; Zdanovich (ed.) 
1995b; 1997; Zdanovich G and D. 1995; Vinogradov 1995a; Malyutina and 
Zdanovich 1995). 

At the settlement of Malokizil'skoe, over an area of 5,000 square meters, 
there was found a ditch, 1.1-1 .4m wide and 1 .4m deep, and part of a shallow rec- 
tangular house with post -holes and hearths/fireplaces. A child cremation burial 
was found in the house, executed victims and the burnt remains of people were 
found both on the square and in the ditch, and sacrificial deposits of cattle and 
pottery were found in the ditch (Sal'nikov 1967: 19-20, 35-38, fig. 3: 1-8). The 
settlement was probably destroyed by fire. Malokizil'skoe is the only settlement 



32 CHAPTER FOUR 

in the "land of towns" that is assigned to the Abashevo culture; all others belong 
to the Sintashta-Petrovka types. 

The fortification and layout of the settlements were deliberately planned in 
advance, taking into account the natural relief. Sites are surrounded by a ditch, 
2.5-5m (Sintashta), 4.5m (Ust'e), and 5-7m (Kuysak) and 1.3m deep, with two 
rows of defensive walls, 1.7m and more thick made of clay blocks and vertically 
erected pine logs 0.4-0. 5m in diameter (Ust'e). Walls were also made of timber 
frameworks filled with earth; there was probably a timber palisade above them. 
The ditch was cut in steps and reinforced by logs. 

The inner square of the fortresses revealed regular planning and was divided 
by radial and perpendicular roads, along which a cart or a chariot could pass. The 
roads led to entrance gates. The houses were situated between the inner and outer 
walls; they were rectangular or trapezoidal in form with sides measuring 20m x 
13m x 17m (Kuysak) or 15.5-20m x 5-7m x 7.9m. Houses abutted the outer wall 
and shared a common roof with a pitch towards the center; entrances faced the 
central square; the house walls were indicated by post-holes. Wells and round 
surface hearths were found inside the houses. There are traces of slag and copper 
beads in every building (Grigor'ev 1994). For metal working, they employed 
furnaces, 0.5 -0.9m in diameter and 0.3 -0.6m deep, that were connected by a 
narrow trench, 0.3m wide and 0.7-1. 2m long, leading to the well (Kuysak). 

There were numerous ritual burials of children, dog, cattle, sheep and goat, 
especially kids' sacrifices; vessels are found in defensive ditches. Traces of 
massive fires and destruction, apparently caused by warfare and demanding the 
rebuilding of the settlement, have been found on many sites. 

The settlement at Kuysak is an irregular rectangle of outer walls, 122 x 96m, 
surrounded by a wide ditch. Inside the site there is an oval-shaped wall 
measuring 64 x 58m. The outer and inner walls are connected by three walls 
forming segments around a central square. In each segment were 6 to 9 houses 
(Malyutina and Zdanovich 1995). 

Partially inundated by the river and covered by constructions of Alakul', 
Fedorovo and Alekseevka types, the Sintashta settlement was originally of 
circular form, 136-147m in diameter, and surrounded by an earthen wall girded 
by a ditch. Habitations were of trapezoidal form that only partially adjoined the 
outer wall. There was probably a second badly preserved wall that surrounded 
the central oval, 60-65m in diameter (Gening et at. 1992). 

The settlement of Arkaim presents a most impressive view (Fig. 58: 5). The 
area of settlement comprised 20,000 square metres and formed two concentric 
walls, composed of clay blocks, enclosing a central square, surrounded by a ring 
of habitations that adjoined the first defensive wall (85m in diameter). There was 
a circular street and outer circle of habitations abutting the more formidable 
outer wall (Zdanovich 1989: 181-182; 1997: 48-50, fig. 3). The diameter of the 
outer wall measured 143-145m and was 3-3. 5m thick. It was composed of wood 
and clay with added lime; it was faced on the outside with clay blocks. There 
was a timber-faced ditch some 1.5-2m deep. The whole construction was divided 
by radial streets, which probably led to the gates. The western gates were 5-6m 
wide, others were labyrinthine. Logs and blocks were used to construct the trape- 
zoidal houses. They consisted of a timber frame filled with soil. The habitation 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 3 3 

area measured 110-180m and 0.4-0.5 deep. Wells and furnaces connected with 
metal working were found in the habitations. 

Numerous traces of metal production have been discovered in all the fortified 
sites of the 17th— 16th centuries BC. That is why we agree with those who have 
suggested that the fortresses were constructed by early metallurgists to guard 
areas of copper deposits. The defensive nature of the forts and the traces of 
burning horizons and reconstruction indicates the unstable situation in the 
steppes of the Sintashta period. 

The origin and chronological correlation of early Andronovo sites are based 
on stratigraphy. On the Kuysak settlement ceramics with Pit-grave/Poltavka and 
Abashevo features were recovered from the earliest levels of the ditch. It 
indicates that these cultures participated in the formation of the Sintashta type of 
site (Smirnov and Kuz'mina 1976; 1977; Gening 1977). The early settlement 
built by the Pit-grave/Poltavka and Abashevo tribes was burnt, then it was rebuilt 
in the Sintashta period only to be fired again. All of this reflects the extremely 
tense atmosphere in inter -ethnic relations (Malyutina and Zdanovich 1995: 104- 
105). 

The stratigraphic correlation of the Sintashta and Petrovka types is indicated 
at the Ust'e settlement where the early Sintashta settlement was circular and the 
houses were trapezoidal. In the Petrovka period the settlement was rebuilt in a 
rectangular form with extended rectangular houses, 160 square meters in area; 
the adjoining walls were divided by a main street (Vinogradov 1995). Similar 
planning has been discovered on Petrovka-type sites in the Urals and in north 
Kazakhstan that chronologically follow Sintashta sites. 

What place do the Sintashta sites occupy in the Eurasian cultures of the Early 
Bronze Age? Large fortified settlements of this period were long known in the 
western steppe. These include Kamenka in the Crimea, and Karataevka and 
Liventsovka in the Don region (Bratchenko 1976: 119-122). The Liventsovka 
fortress is a semicircular promontory fort, 20-24m high, enclosed by a double 
semi -circle of massive stone walls and surrounded by ditches, 2-6m wide and 2- 
3m deep. The walls consisted of an inner and outer course of large stones with 
the inner space filled with small crushed stones. The habitations were located 
between the two walls as in Arkaim and were also of trapezoidal form. A large 
quantity of (presumably enemy) arrowheads was found outside the fort. The 
Crimean and Don forts are close to the Multi-roller Ware culture that 
chronologically follows the Catacomb culture. 

Defensive constructions are also associated with the Abashevo settlement 
(Pryakhin 1976: 23-24). It possesed a ditch, 2.2-2. 8m wide and 0.4-0. 5m deep. It 
enclosed a palisade indicated by a double row of post -holes, standing 2m from 
each other in a row; the width of the main entrance was 4m, the gate to the river 
was 1.5m. 

Further west defensive banks and ditches are well known on Bronze Age 
settlements of the Danube area in the Monteoru and Filzesabony cultures 
(Mongayt 1974: 68, 86). In the Balkans we have circular fortified settlements 
from the end of the 4th to the beginning of the 3rd millennium BC at sites such 
as Ploskata Mogila and Ezero. It is possible that the spread of this type of 
monument can be traced back to Anatolia where it originated already in the 6th- 
5th millennia BC (Hacilar) through Troy which has been culturally tied with the 



34 CHAPTER FOUR 

Balkans. N. Ya. Merpert (1995) has examined the Balkan sites and regarded this 
hypothesis questionable. He suggested the "polycentric character in the develop- 
ment of this architectural tradition" which could have appeared in the Balkans 
under the influence of central and east European cultures, including the Tripol'e 
culture. I might suggest that the formation of fortified settlements in the Eurasian 
steppes was based on the principle of the defense of a military camp. We might 
look for its origins in the concept of the circle camp of kibitkas, consisting of 
covered wagons (with their rear-ends facing outside) forming a shelter for the 
cattle, women and children in the center. Such an idea could only arise on the 
steppes, among people travelling in wagons. Such defensive tactics was used by 
steppe peoples over the millennia: it is attributed to the Vedic Aryas (Rau 1983) 
and later it is found among Turks, Mongols and Cossacks till the 19th century 
AD. 

Both Liventsovka and Arkaim exhibit spatial plans that resemble that of the 
Koi-Krylgan-kala temple complex of the 5th century BC in Khorezm that recalls 
an idea of S. P. Tolstov (1948: 77-82) about "towns with inhabited walls," which 
he compared to the idealized settlement of the ancient Iranians described in the 
A vesta, the vara, which was built by their legendary ancestor and first king, 
Yima. 

Yima Xsaeta kneaded "clay with his feet, divided it into pieces with his 
hands... and made the Yima vara, a horse run long (nearly 800m) on all four 
sides. He brought seed of sheep and goats, horses, dogs, birds, fires, red and 
burning. And made Yima this vara... house for people... shed for cattle. He led 
water there along the route for the length of a haOra; he built a house there, vault, 
yard-place closed from all sides" (Videvdat 2.33 and further). 

V. A. Livshits (1963: 145) regarded the vara as a fortified settlement to 
shelter people and cattle during wartime. 

There was also the Old Persian term dida which M. M. D'yakonov (1961: 67, 
365), following E. Herzfeld, derived from "(knead) clay". This word denoted a 
defensive wall, the fortified settlement of a clan. The Vedic Aryas knew several 
terms to denote a fortified place: fortress, rampart (Elizarenkova and Toporov 
1995: 512). However, even in the Achaemenid period, the Persians did not know 
of the town proper: Pasargadae and Persepolis were only residences and cult 
centers. 

In this respect W. Rau's interpretation (1983: 11) of the Sanskrit term grama 
is of major importance: later it denoted village, but initially it denoted a small 
tribal group of shepherds migrating in wagons. Wagons were put in a circle 
every evening, forming a peculiar fortress on wheels, with the cattle inside 
(Elizarenkova and Toporov 1995: 490). 

K. Jettmar (1981) and V. Brentjes (1981) interpreted the Bactrian plans of 
cult centers such as Dashly III and Dzharkutan, discovered by V. I, Sarianidi 
(1977) and A. Askarov (1977), as models of a vara. The latter, dated to the 14th- 
13th centuries BC, have neither sources nor parallels in the Near East, and they 
repeat the architectural plans found at Arkaim, which presupposes the influence 
of the northern steppe culture on the architecture of Bactrian farmers. This con- 
clusion is supported by Andronovo ceramics and decoration on temple vessels 
found on a sacred altar of the Dzharkutan temple (Askarov 1989). 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 3 5 

Of great importance is also the fact that the same planning principle was 
preserved in the steppe during the Scythian period in the construction of the 
royal kurgans, e.g. Arzhan (Gryaznov 1980) and Tagisken (Tolstov 1962: 21, 81- 
86). This prompted K. Jettmar (1981) and L. A. Lelekov (1972) to return to the 
problem of the Avestan vara. The vara was not an actual town but, as the late 
Avestan Videvdat text would have it — a model of the universe or micro-cosmos. 
But the mythologeme itself could only have originated in a society where there 
already existed the prototype of a 'celestial town', i.e., in the culture of the 
population of the Eurasian steppes at the beginning of the 2nd millennium BC. 

I. V. P'yankov (1999) supported K. Jettmar's comparisons of the vara with 
the Bactrian cult centers and like him regarded Arkaim as a cult center. But this 
is unlikely as the architectural plan of Arkaim is typical of the other Sintashta 
monuments that indicate not only a ritual but also a production center. He 
remarked that the Yima-Imra cult occupied a central place in the mythology of 
the third group of Indo-Iranian peoples, the Nuristani, who preserved 
recollections of the vara-type construction. 

I. M. Steblin-Kamensky undertook very interesting research into the vara 
problem (1955: 166-167). According to him, the vara, which was erected in the 
legendary homeland of the Iranians in the ariiansm vaejo by the first mortal and 
the first shepherd-king of Iranian mythology Yima, had a round and not a rectan- 
gular form. It was oriented "to all four sides of the world" and it consisted of 
three concentric circles of clay walls with 9, 6 and 3 passages. The vara plan 
thus corresponds to the round settlements of the Arkaim type. 

Of considerable importance here is I. M. Steblin-Kamensky' s observations 
about the borrowing of the cult of Yima, who is recalled in Indie mythology as 
Yama, King of the Dead, by Finno-Ugrian peoples. The latter adopted the cult of 
the god Jumula in Finnish or Joma in Komi, the word vara (cf Hungarian var 
'fortress' and varos 'town'), and the two golden symbols of power presented to 
Yima by Ahura-Mazdah: a shepherd's horn and a stick with sharpened end for 
goading animals. The Shepherd-king expands the lands of the Aryans to the 
south, which "can indicate the direction of Aryan migration". 

It seems significant that already in the Sintashta and Petrovka periods the 
ditches surrounding settlements were used for ritual purposes: on their bottoms 
archaeologists have recovered sacrificial complexes in the form of several 
vessels, and the skeletons and bones of animals. This sacral function 
consequently may have carried over as the circle and square components of a 
temple and royal burials which would have been constructed as microcosms of 
the universe. 

The fortifications of Sintashta make it possible to return to the question of 
forts among the Vedic Aryans. There were two terms in the Indo-European 
languages to denote fortified settlement: 1) burg lost in Indo-Iranian, Tocharian 
and Latin, and 2) from the Aryan-Greek-Baltic area, Greek polls, Lithuanian 
pills and early Indian pur, ultimately derived from a root 'precipice, steep slope' 
(Gamkrelidze and Ivanov 1984: 744-746). M. Wheeler (1984), S. Piggott (1950: 
261) and many researchers after them have identified the pur as the Harappan 
fortresses, destroyed by Indo-Aryans and, once reduced to ruins, they became the 
arma or armaka, where Aryans gathered ceramics for fire-clay necessary for the 
manufacture of Aryan utensils. W. Rau (1983) suggested that the pur was not a 



36 CHAPTER FOUR 

fortress of aboriginal enemies but a small settlement of Aryans proper, enclosed 
within an earthen or stone wall to defend cattle, and the anna denoted a deserted 
Aryan site. However, in the numerous Rigvedic texts (IV, 30,20) analyzed by A. 
Parpola (1988: 208-212) Indra, the main god of the pantheon, has the recurrent 
epithets "destroyer of puf and "victor over the Dasas". Dasa implies both 
demons and the dark-skinned aboriginal population (Bongard-Levin and Gurov 
1988; 1990; Elizarenkova 1989: 433; Alekseev 1990; Bailey 1959: 107-115). 
The Rigveda (2.20.7) directly mentions the destruction of Dasa fortresses; more- 
over, their kings do not appear to have Indo-European names (Parpola 1988: 
212). Indra destroys 99, 100 and 101 Dasa towns; Agni burns them with fire. 
Despite the opinion of a number of scholars that there was a chronological break 
between the collapse of the Harappan towns and the arrival of the Indo- Aryans, 
these texts demonstrate that the Aryans did participate in the destruction of the 
towns of the Indus Civilization. At the same time the term pur itself, common in 
Greek, was also in use in the Indo-Iranian homeland where it would have been 
applied to fortified settlements of the Liventsovka and Arkaim type (in this 
respect the ideas of S. S. Berezanskaya (1971) on similarities between the stone 
architecture of the North Pontic area and Achaean Greece are of interest). 
However, by the mid 2nd millennium BC, the tradition of constructing 
fortifications declined due to the stabilized situation on the Eurasian steppes. 
Pastoral tribes did not follow the path toward urban development that had begun 
in the 17th- 16th centuries BC but rather turned to the extensive development of 
their economy by migrating and mastering new territories. 



House-plan 

The settlements and houses of the later stages of Andronovo cultural 
development have been rather well studied. Houses have been discovered on 
more than 200 settlements in different regions, some of which have been ex- 
cavated. The most fully studied are Spasskoe, Alekseevka, Sadchikovo, Kipel', 
Zamaraevo, Sarafanovo, Duvanskoe 17, Uk 3, Mirny 3, 4, Tasty-Butak, Ushkatta 
1, 2, 8, Shandasha, Yazevo 1, Kamyshnoe 1, 2, Pereleski, Zagarino, Semi- 
ozernoe, Chaglinka, Bishkul' 4, Bogolyubovo 1, Petrovka 2, 4, Novonikol'skoe 
1, Pavlovka, Yavlenka, Sargary, Atasu, Buguly 1, 2, Akbaur (Akkauzen), Myr- 
zhik, Shortandy-Bulak, Tagibay-Bulak, Karkaralinsk, Ikpen' 1, Entuziast, Mayo- 
rovka, Kent, Dongal, Tashik, Kanay, Trushnikovo, Malokrasnoyarka, Shlyapova 
(Krivtsova-Grakova 1948: 73-100; 1951: 153-177; Sal'nikov 1951: 102-105; 
1954a; 1957; 1959a; 1967: 242-248; Sal'nikov and Novichenko 1962; Margulan 
1959: 16-27; Margulan 1979: 163-254, fig. 119, 124, 128, 130, 132, 137, 143 
146, 154, 167, 169-171, 185; Margulan et al. 1966: 197-255; Chernikov 1960 
26-35, 40-44; Sorokin 1962a: 51-60; 1966: 61; Kuz'mina 1962a: 88-91; 1962b 
9-11, fig. 1; 1964b: 101-104, fig. 29; 1986b: 42-55; Fomina 1964: 207 
Zdanovich 1970: 147; 1973a: 113-118; 1973b: 40-51; 1973c: 21-43; 1974a: 61: 
1974b: 61-62; 1975; 1980c; 1982; 1983; 1987; 1988: 19-60; Zaybert et al. 1974 
71-75; Evdokimov 1971: 65-67; Chemyakin 1974: 50-55; 1972: 288-289; 1978 
201; Chebakova 1975: 92-97; Potemkina 1976a: 10-16, 23; 1982: 15-49; 1983 
1985: 30-145; Gusentseva et al. 1972: 204; Zdanovich S. 1978; 1979; Orazbaev 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 



37 



1970: 129-146; 1972: 109-114; Kadyrbaev 1983; Kadyrbaev et al. 1992: 146 
Korochkova 1984; Korochkova and Stefanov 1983; Koryakova et al. 1991 
Malyutina 1990; 1991; Varfolomeev 1987; 1988; 1991; Tkachev 1987; 1989 
1991; Kurmankulov 1988). 



Area 



Settlement 



Number of houses 



The Urals Bakhtinskoe 

Low Bakhtinskoe 

Low Spasskoe 

Zamaraevo 

Tribiyak 

Barmino 

Chernigovskoe 

Bely Kamen' 

Talovskoe 

Kipel' 

Konezavod 

Pereleski 

Kulevchi 3 
Northern Kazakhstan 

Sargary 

Chaglinka 

Petrovka 2 

Novonikol'skoe 

Atasu 

Obaly 

Shandasha 

Ushkatta 2 

Tasty-Butak 
Western Kazakhstan 

Atasu 2 (Ak-Mustafa) 

Atasu 1 

Buguly 1 

Buguly 2 

Shortandy-Bulak 

Suukbulak 

Tagibay-Bulak 

Aynakol' 

Sorkuduk 

Myrzhik 



7 

10 

10 

10 

13 

15 

17 

19 

20 

20 

6 

10 

5 

13 
16 
16 
15-20 

22 

24 

2 

2 

6 

49 

23 

100 

30 

38 

12 

6 

8 

18 

37 



Table 1 : Number of houses per Andronovo settlement 

The table is selective rather than comprehensive as there are no data about many 

settlements in the literature, and many have been partially destroyed. 



Andronovo settlements were usually situated on the banks of small rivers, on 
grounds defended from wind by rows of hills or riverside rocks. In the Urals they 
are connected with river valleys with large floodplains; very often they are in the 



38 CHAPTER FOUR 

very floodplains inundated in modern times. In west, north and central Kazakh- 
stan settlements are usually situated on the first river terrace. 

Apart from permanent settlements temporary sites appear in the Final Bronze 
Age near mines (Mungly, Akzhal, Milykuduk) or on desert pastures far from 
rivers, where they were dependent on artificial wells. 

Andronovo settlements are divided into two types according to area: small 
settlements with only a few houses and large settlements with 10-20 structures. 
Small settlements with only 2-6 houses are typical of the western region (Shan- 
dasha, Ushkatta 2, Tasty-Butak). They resemble neighboring Timber-grave and 
Tazabagyab settlements. In the Urals and in north and central Kazakhstan the 
large settlement of 10-20 houses predominates (Table 1). In the Final Bronze 
Age even larger settlements with several dozens of structures are known (Shor- 
tandy-Bulak, Buguly 1,2, Akbaur, Kent, Myrzhik). 

Viewed chronologically, we can observe the dynamics of Andronovo house 
types, tracing them through the Bronze Age as they gradually adjusted their 
houses to local conditions and obtained the skills that allowed them in the Final 
Bronze Age to pass over to nomadic cattle-breeding and master the deserts and 
mountains of Central Asia. 

Settlements of the Petrovka type of the 16th century BC derived directly from 
the Sintashta type, with which they were genetically connected. In the Urals and 
northern Kazakhstan we find traces of fortifications on Petrovka sites: Petrovka 
2, Novonikol'skoe 1, Bogolyubovo 1, Kulevchi 3, Semiozernoe, Konezavod and 
the Andronovo settlement of Chernoozerje on the Irtysh that yielded Fedorovo 
ceramics. Archaeologists have uncovered ash-filled ditches, 1.5-2. 5m deep, 1.2- 
3.5m wide, with inner and outer banks measuring 0.4m high and passages 2- 
2.5m wide. At Bogolyubovo and Semiozer'e the ditch cut the promontory from 
the field side; in Novonikol'skoe it enclosed a rectangular area 95 x 60m; in 
Petrovka 2 the inner ditch divided an inner area, 70x120m, into two parts; in 
Chernoozer'e the ditch, 75m long, adjoined a high river terrace enclosing a 
rectangle on three sides. Post-holes have been discovered along the outer wall 
probably from a palisade in Chernoozer'e (Zdanovich 1975: 7-10; 1988: 133 
Zdanovich and Gening 1985: 151; Viktorova and Borzunov 1974: 19-20, fig. 1 
Potemkina 1982: 52, fig. 1; Vinogradov 1982: 97). 

According to the evidence from northern Kazakhstan, there the houses were 
arranged randomly; they were small, 40-50 square meters in area, surface or 
semi-subterranean (0.15-0. 3m deep, rarely 0.6-0. 8m); they were of post-built 
construction (in every house 20-30 post-holes are found), and they had hearths 
from 0.5-1 to 1.5m in diameter and 0.2-0. 5m deep, sometimes faced with stone 
and the floor coated with clay. Ash-pits adjoined the houses (Zdanovich 1975: 9, 
21; 1976b: 63; 1973a: 115; 1988: 19-60; Vinogradov and Zdanovich 1979: 161- 
162; 1980: 138; Vinogradov 1982: 94-99). 

Continuity from the architecture of Sintashta is seen in both the tradition of 
fortification and in the predominance of slightly subterranean houses. But the 
Petrovka settlements already reflect a somewhat different ethno-political and 
economic situation. 

The next stage in the development of Andronovo house building is seen in 
the settlements of the mature Bronze Age. In architectural form, the houses are 
genetically connected with the Petrovka structures. In the Orsk region and in the 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 3 9 

Urals a large number of settlements with a pure Alakul' or Kozhumberdy layer 
has been discovered (among them 20 settlements are in the Elenovka micro- 
district, including Elenovka 1, 2, Kiimbay 1,2, 3, Ushkatta 1,2, 8, Shandasha, 
Tasty-Butak, Tanalyk Starikovskoe, Kamyshnoe 1, Kambulat, etc. 

Settlements with a frequent Fedorovo layer are known in the Urals 
(Duvanskoe 17), in northern Kazakhstan (Bishkul' 4, Pavlovka), in central 
Kazakhstan (Ikpen' 1, middle layer) and in the east of the area, in west Siberia 
(Shlyapova, Klyuchi, Bateni). 

A pure layer from the Final Bronze Age is seen in northern Kazakhstan in 
Chaglinka, Sargary (Fig. 86: 1-9), in eastern Kazakhstan in Malokrasnoyarka and 
Trushnikovo, in central Kazakhstan in Shortandy-Bulak Karkaralinsk, Tagibay- 
Bulak, Suukbulak, Kent, and Myrzhik. A majority of the settlements has a mixed 
cultural layer, which points to continual re-occupation of the same site 
(Alekseevka, Sadchikovo, Kipel', Zamaraevo, Novo-Burino, Chernorech'e, 
Yazevo, Kamyshnoe, Petrovka, Novonikol'skoe, Yavlenka, Atasu, Buguly 1, 2, 
etc.; Fig. 86). 

As a rule Andronovo settlements are of regular construction, houses are 
oriented similarly and they are situated parallel to each other, taking into account 
the topographic conditions: entrances face the river or are on the leeward side 
(Fig. 7). There are three types of settlement plans: 

Type I settlements find their houses situated in a single line along the river 
(Ushkatta 1-8, Zamaraevo, Chaglinka, Konezavod, Sorkuduk, Karkaralinsk 2, 
Tagibay-Bulak). 

Type II settlements have houses that are built along the river in two parallel 
rows, divided by a street 7-10m wide (Alekseevka, Sadchikovo, Shandasha, 
Atasu 3, Marzhan, Myrzhik). 

Type III settlements date to the Final Bronze Age and are found in central 
and northern Kazakhstan. The settlement, extended along a river, is almost rect- 
angular or oval in plan with a large square in the center which was devoid of 
buildings; it served for keeping cattle (Zhabay-Karasu, Novonikol'skoe 1, Vino- 
gradovka 6, Shortandy-Bulak, Akbaur 2, Buguly 1) (Margulan et al. 1966, fig. 
100-102, 105; Margulan 1979, fig. 137, 145, 149; Zdanovich S. 1979: 8). Some- 
times houses are situated close to each other over the whole settlement area 
(Atasu 1, Sargary, Kent, Dongal). 

In Alekseevka, Sadchikovo, Atasu, and Akbaur there are traces of wooden 
palisades around the settlement or fences to enclose part of the settlement like 
Kazakh boskets made of wood, earth or stone which serve to guide cattle or 
protect them from wind and snow (Krivtsova-Grakova 1948: 97-99; Margulan 
1959: 19). 

The existence of large ash-pits, 0.5-lm and more deep, associated with a 
domestic hearth cult, is a particular feature of Andronovo settlements. Three 
types of ash-pits have been recorded: 1) large deep pits adjoining a house wall 
(Sadchikovo, Zamaraevo, Mirny 4, Atasu, Malokrasnoyarka); 2) an elongated 
trench along a river bank or edge of settlement (Kamyshnoe 2, Obryv, Ushkatta 
2, Shandasha, Mirny 4, Kanay); 3) a mound adjoining the house (Alekseevka, 
Obaly, Chaglinka, Atasu). This type may probably be considered the most 
recent, serving as a transition to the ash-pits of the Scythian period (Grakov 



40 CHAPTER FOUR 

1971). In addition, the foundation slots of abandoned houses (Alekseevka) or 
trenches (Novonikol'skoe) were used for ash-pits. 

There are two categories of buildings known from Andronovo settlements: 
semi-subterranian and surface structures. 



House Type I 

The first category of Andronovo house is long-term semi-subterranean structure, 
set 0.6-lm (rarely 1.2-1. 6m) deep in the earth (Fig. 8). 

The construction technique of the Andronovans is revealed through the study 
of both houses and burials. Grave and house foundation ditches were dug with a 
pick-ax or adze with a narrow blade (8-13. 5cm); traces of tool marks are found 
on the walls of a house at Sadchikovo and the Uvak grave. Earth thrown up from 
the foundation ditch was placed outside to form a breastwork which raised the 
height of the house approximately lm. A mixture of clay, sand and stones was 
used in central Kazakhstan (Margulan 1959: 21, 22). In other regions the breast- 
work was covered by a layer of clay (Kuz'mina 1962; Malyutina 1990). 

Settlement Size (meters) 

Urals 

Duvanskoe XVII 10.5x5 

Kipel' 11x10 

SpasskiyMost 14x6 
Verkhnealekseevskoe 12x7 

Bely Kamen' 30 x 20, 1 1 x 5 

Bakhtinskoe 27x16,19x11 

Zamaraevo 26 x 1 1 

Alekseevka 15x9,19x16,26x15 

Sadchikovo 17 x 11 

Mirny III 18.3x9 

Mirny IV 12x8.5,13x12 

Kambulat 27 x 10-23 

Zagarinka 24.8 x 12 

Kamyshnoel 12.8x7.5, 11.5x7.5, 12.0x10.0 

Kulevchilll 24.5x8.5-10.5, 14x4-5.5,4x6.2,14-15x11-12, 12x7-8.5 

Yazevo I 16.5 x 12.0, 14.6 x 7.2, 14.5 x 12.5, 14.0 x 8.0 

North Kazakhstan 
Chaglinka 15x5,10x8,16x14,30x10 

Il'inka 23 x 12 

Novonikol'skoe I 22.0 x 12.0, 25.0 x 10.0-12.0, 19.0 x 10.0-12.0, 18.0 x 13.0, 

17.0 x 13.0, 14.0 x 6.0-8.0, 13.0 x 10.0, 14.0-8.0 
Petrovka II 24 x 8.5, 20.0 x 9.0, 26.5 x 10-12.5, 12.0 x 12.0, 16.0 x 13.0(?) 

PetrovkalV 13.5x12,16.5x15.0 

Pavlovka 18,0x11.5,16.0x10.0 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 



41 





Western Kazakhstan 


Ushkatta I 


20x10 


Ushkatta II 


25x9 


Ushkatta III 


20x8 


Ushkatta VIE 


16x16 


Shandasha 


11x7 


Tasty-Butak 


18x8,20x10 




Central Kazakhstan 


Atasu 


10x8, 12x11, 13x12,15x11,20x15, 18x12 


Tagibay-Bulak 


from 88 to 375 sq m. 


Shortandy-Bulak 


16x10, 15x10,25x15,20x20 


Suukbulak 


14x8 


Karkaralinsk 


20x12, 18x4 


Baybala 


11x10 


Akbaur 


22x12,30x11 


Buguly 


32x7 


Sorkuduk 


12x8 


Kulman 


12x8 


Kent 


6.5 x 5.7, 7.3 x 5.5, 5.0 x 6.2, 7.5 x 6.8, 7.5 x 6.5 


Domalaktas 


14.4x9.0 


Dongal 


10.2 x 7.2, 10.1 x 8.0, 1 1.7 x 5.7, 9.3 x 8.4 


Tashik 


15.5x9.0 


Upais 


10.2x7.5 


Ikpen' 1 


11.0x10.0,24.0x9.5-12.5 


Ikpen' 2 


18.0x7.0 



Table 2: House dimensions 

The table has been compiled selectively. Not all houses excavated have data published. 

Data of special workshops of Central Kazakhstan are not included in the table. 



Andronovo houses are usually rectangular in form and are distinguished by their 
large dimensions (areas from 80-100 square meters to 200-300 square meters; 
Table 2). A frequent peculiarity of Andronovo house building is the tradition of 
connecting two neighboring buildings with an underground passage, resulting in 
what would appear to be a figure-of-eight structure on the surface. In the Final 
Bronze Age the underground passage system could unite several neighboring 
structures in a single block (Atasu, Sargary, Petrovka 3, Pereleski, Chaglinka). 
The houses of the Alakul' and Kozhumberdy groups often exhibit various out- 
houses which produce their own complex configuration (Kambulat, Ushkatta). 

The pinacle of house -building technique is found in the Final Bronze Age of 
central Kazakhstan. Here we find a type of strictly planned, multi-chambered 
dwelling, rectangular in form and built from massive, well-polished stone slabs. 
Such multi-chambered houses, uniting 5-7 rooms, have been discovered at 
Buguly I, II, III and Akbaur (Margulan 1959: fig. 10, 11; 1979: 187). Domestic 
and industrial buildings are connected with the house, thus, the whole complex 
of structure 17 at Buguly covers 530 square meters; house 22 occupies 660 
square meters and house 28 reaches an extraordinary 1500 square meters. These 



42 CHAPTER FOUR 

are the largest houses known in the Eurasian steppes during the Bronze Age of 
this region (Margulan 1979: 187-189, fig. 140-142). The main house and house- 
hold annexes are on a single axis or at a right angle to one another. 

Andronovo semi -subterranean houses are divided into two types in terms of 
construction: type 1) with wooden constructions; type 2) with stone. The first 
type is represented in the forest-steppe area of the Urals, northern Kazakhstan 
and in the heavily forested regions of central and eastern Kazakhstan. Houses 
here have a timber-frame construction with posts along their walls. These held 
horizontally fitted rows of beams or blocks (Kipel', Zamaraevo, Chaglinka, Pere- 
leski, Kulevchi 3, Yavlenka 1, Petrovka 1, Pavlovka, Ikpen' 1, etc.) or wattling 
covered with clay (Alekseevka, Chaglinka, Kanay). There is a large number of 
post-holes in the center of the house that supported the roof; there are also inner 
partitions, wooden flooring and tables. Within a single house there might have 
been anywhere from several dozens to 200 or even 300 posts. Similar post con- 
structions are known at Alekseevka, Sadchikovo, Kipel', Zamaraevo, Mirny IV, 
Zagarino, Pereleski, Yazevo, Kamyshnoe I, II, Petrovka II, Il'inka, Yavlenka, 
Bishkul' IV, Pavlovka, Chaglinka, Novonikol'skoe, Karkaralinsk, Suukbulak, 
Shortandy-Bulak, Malokrasnoyarka, etc. In some settlements there is evidence 
for only a few or no post-holes; here there was probably a timber framed 
construction (Spasskiy Most, Kambulat, Novoburino). 

The Andronovans employed as building materials birch, pine and cedar 
(Siberian pine), rarely other species. There are sharp axe marks on logs which 
measure 0.25m, sometimes 0.4m and even 0.8m in across. Judging from the 
wooden constructions of Andronovo houses and especially graves (Sal'nikov 
1951b; 1952b; Kuz'mina 1973a) which were, in effect, miniature versions of 
houses, the Andronovans were familiar with: 1) false framework - a layer of logs 
set on each other in several rows of beams without fastening the corners; 2) walls 
built from whole logs or blocks split in half that were turned with their flat sides 
to the interior and fastened to the walls by post -uprights set at certain intervals; 
this was the commonest system of construction. Sometimes a layer of logs was 
set between two rows of poles, corners are butt -joined (Fedorovka, Sosnovskiy, 
Verkhneozernoe, Borovoe); 3) timber-frames with bond jointing: in the lower 
timber rank the longitudinal logs are long, and then butt-ends are short; in the 
next timber rank the longitudinal logs are short and the butt-ends are long 
(Tuktubaevo); 4) timber-frames where the logs are joined by inserting one log 
into a groove of the lower log (= American Lincoln log system); 5) timber- 
frames where the upper log is joined to the lower by mortise and tenon 
(Tuktubaevo, Fedorovo). 

There was a variety of ways of constructing a roof or cover from poles, logs 
or blocks placed along the longitudinal axis of the pit, across it, resting on a 
horizontal log placed atop a post inserted into the center of the grave (Sintashta I, 
II, Sineglazovo, Tursumbay) or on two or four posts on the corners of the pit 
(Sintashta II, Sosnovskiy, Verkhneozernoe, Fedorovo); there are also overhead 
covers consisting of two layers laid perpendicular to each other (Alakul', 
Kinzerskiy). 

Semi-subterranean houses of Type 2 are found in treeless districts where 
stone was used in architecture. We can divide these into various sub-types. Sub- 
type IIA consists of large stone slabs, 0.3-0. 5m wide, 0.75-1. 4m long, sometimes 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 43 

reaching 1.55 x 2m, that were set vertically on edge to a depth of 0.2-0. 4m. The 
slabs were stacked dry without mortar; they were often placed in two rows 'in 
bond'. Natural outcrops of flagstone near settlements served as quarries. Houses 
with stone walls have been examined in the Orsk region (Ushkatta I, II, VIII, 
Shandasha, Tasty-Butak) (Kuz'mina 1962a: 88-91; 1962b: 9-11, fig. 1; 1964b: 
207, fig. 29; Sorokin 1962a: 51-60) and in central Kazakhstan (Atasu, Buguly, 
Akbaur, Baybala, Sorkuduk, Karatomar, Tagibay-Bulak). On the Final Bronze 
Age settlements of Buguly II, Akbaur and Kent we find sub-type IIB: the use of 
a stone cist construction, sometimes employing clay mortar. In Tagibay-Bulak 
the space between the two rows of slabs was filled with rubble (Margulan 1959: 
16-27; 1979: 163-254, fig. 137, 140-143, 146-8, 169-171; Varfolomeev 1991: 7). 

The presence of stone or wooden architecture depended solely on local 
building resources. 

Sometimes there are post -holes from the uprights that supported the roof in 
the center of a house with stone walls (Atasu, house 17). In other cases post- 
holes have not been found (Spasskiy Most; Novoburino, house 2; Tasty-Butak, 
Shandasha, Atasu, house 20) or a house may reveal only several small post -holes 
without any discernable pattern (Ushkatta, Kambulat) which suggests the 
erection of a roof covering without posts. 

The evidence suggests several types of house roof (Fig. 9): 

A roof with two sloping surfaces supported by a longitudinal log (ridge) lying 
on supporting posts inserted along the center of the longitudinal axis of the house 
(Zamaraevo, Buguly 2, Suukbulak) (Sal'nikov 1954a: 246; Margulan et al. 1966: 
253, fig. 113). A parallel to this can be seen in the Timber-grave settlement at 
Moechnoe Ozero (Merpert 1958: 117-118; Trubnikova 1958: 186-187, fig. 3; 
Sinitsyn 1949: 199). 

A roof with two sloping surfaces supported by two rows of posts erected 
along the longitudinal axis of the house and covered by two beams supported by 
cross beams (Malokrasnoyarka) (Chernikov 1960: 41). The Timber-grave settle- 
ment at Suskan offers parallels to this technique (Sal'nikov 1952b: fig. 19; 
Merpert 1958: 109-110, fig. 13). 

A roof consisting of four sloping surfaces supported by a square frame in the 
center of the house; the frame was supported by four posts inserted around a 
central hearth with a smoke -hole for draught (Alekseevka, Kipel', Chaglinka, 
Shortandy-Bulak, Suukbulak) (Krivtsova-Grakova 1948: 77-81; Sal'nikov 
1951b: 107; 1954a: 246; Orazbaev 1970: 142; Margulan et al. 1966: fig. 126; 
Margulan 1979: 202, fig. 152). 

Pyramidal-stepped vault. The existence of this type is indicated by the 
discovery of rectangular frames measuring 1.8 x 1.3m and a burnt covering at 
Bishkul' IV and Pavlovka; it is also seen in stone from a house at Shandasha in 
which the roofing slabs were placed along the walls with a lap joint towards the 
center (Kuz'mina 1964: 104; Zdanovich 1988: 22; Malyutina 1990: 111, fig. 2). 
This method was developed in the Final Bronze Age in the cemeteries of Begazy 
and Sangru (Margulan 1979: table 15, fig. 70, 71, 76, 94). 

Reconstructions of how the stepped vault was erected have varied. V. S. 
Sorokin (1962a: 53, 55) suggested that there was a tent-like roof at Tasty-Butak, 
analogous to the one reconstructed for Lyapichev Khutor by M. P. Gryaznov 
(1953: 144, fig. 63): here the house area was divided into squares, each square 



44 CHAPTER FOUR 

covered by a pyramidal log frame -work, laid parallel to the house walls with a 
hole left above the hearth. A. Kh. Margulan (1959: 24, table 4), on the basis of 
the layout of post-holes at Atasu, argued that the house was divided by two 
longitudinal rows of posts into two bays; side naves were covered by a cross 
layer and the central nave was covered by a pyramidal-stepped construction built 
from rectangular log frames. The author (Kuz'mina 1964b: 104) has suggested 
that the post -less houses from the Shandasha settlement could have been covered 
according to the chor-hona system: beams forming the stepped framework of the 
vault were set at diagonal angles, then the beams of the next layer were again 
laid diagonally on a rectangular or hexagonal frame forming a square for the next 
row and so on; a hole was left above the hearth. This system greatly economized 
on the use of wood and such a light covering was widely used in Central Asian 
house architecture and in cult constructions in Kazakhstan (Voronina 1951b; 
Pisarchik 1954: 271-273, fig. 23; Andreev 1958: 436-441; Basenov 1959: 97- 
101, fig. 13, 19). 

O. A. Krivtsova-Grakova (1948: 81) and A. M. Orazbaev (1970: 138) 
suggested that in Alekseevka and Chaglinka the inhabitants used a conical vault 
similar to that of a yurt. The framework of the roof was covered everywhere with 
baira — a layer of reeds (found on the settlement of Kanay) or with layers of 
poles covered with clay (Zamaraevo, Pereleski, Atasu, Kanay, Ushkatta 2, 
Pavlovka) and they placed soil or ash on top which is typical of the Timber- 
grave culture (Krivtsova-Grakova 1955: 75, 100). 

The entrance to the Andronovo house was from the leeward rear-end side; 
sometimes it faced the river and formed a narrow corridor, l-2m from the inside 
to 0.8m from the outside, 2-3 m long (rarely as much as 5m), which descended to 
the floor of the house like a pandus, sometimes with several steps and a 
threshold 0.1 5-0. 2m high (Shandasha, Ushkatta). The steps and threshold at 
Shortandy-Bulak were covered with slabs. A step to which the door was fastened 
was found at Alekseevka. At Petrovka 2 three post-pits from the door 
construction were discovered. In the forest regions the post-holes of the corridor 
posts could be traced; in the steppelands the entrance was formed by strong 
polished tetrahedral pylons (Atasu, Ushkatta II, Shandasha). The house often had 
from one to four additional exits into corridors, 0.8-1. 5m wide and from 2 to 
12m long, that led to neighboring houses or ash-pits. 

The floor of an Andronovo house was earthen with a cup-like depression in 
the center as a result of trampling. In some cases, especially when the ground 
was sandy, the floor would be covered with clay, 0.05 — 0.3m thick (Atasu, 
Karkaralinsk, Novonikol'skoe 1, Ikpen', Dongal, Shandasha), and sandstone 
slabs (Sorkuduk) and probably with wood flooring (Alekseevka, Zamaraevo, 
Petrovka IV). 

The interior plan of the Andronovo house has been insufficiently studied. In 
some cases the house was divided into two sections by a partition wall formed by 
left-over earth; one section might have its floor deeper and only in this place was 
there a hearth. Gryaznov suggested that this compartment was for people to live 
in while the other part was for keeping cattle in winter time (Gryaznov 1953: 
144,147). Division of the house into male and female sections is known in 
Iranian ethnography. Sometimes a space would be divided into compartments by 
partitions (Ushkatta 2, Shortandy-Bulak). The Pavlovka excavation uncovered 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 45 

slots with traces of upright posts that formed squares, the central one measuring 
9 x 9m (Malyutina 1991: 151). 

Andronovo houses were heated and illuminated by hearths of various types. 
Type 1 had an open round or oval fire-place, 0.7-3m in diameter, sometimes 
floored with stone. Such fires, that met both domestic and industrial needs, are 
found both inside and outside of houses. This type of hearth {gulkhan alou) is 
known in Central Asia, first of all among the Iranian Tadzhiks and Pamiri, and is 
to be found in communal houses for men where it originated from early Iranian 
houses of fire (Pisarchik 1982: 72). Type 2 hearths comprise a shallow round or 
oval pit, 0.5-0. 8m in diameter, sometimes more, 0.1 5-0. 4m deep, and often 
covered with flat stone slabs on the bottom (Fig. 10). This is the most 
widespread type of hearth and served for cooking, heating and lighting; it is 
similar to the Central Asian type of hearth known as the chakhlak or chagdon 
(Pisarchik 1982: 78, 79, 109 ). 

This hearth is described in ancient Indian texts as the domestic fire 
garhapatya- 'fire of the master of the house' (Mandel'shtam 1968: 126). Such 
hearths were used for ritual purposes: a bride would go around it, a widow would 
perform a funeral dance, people jumped over it during a feast. The gulkhan 
(hearth) from gul- 'heat' is preserved in the Iranian and Indian languages 
(Pisarchik 1982: 74-77, 105, 106). 

The third type of hearth has a rectangular form, from 0.7 x lm to 1.5 x 2m, 
and was made of closely adjusted rectangular stone slabs inserted into the ground 
on their narrow ends. Such hearths were found in the center of a house, kept 
clean, and it is likely that they had a ritual function (Atasu, Buguly, Shandasha, 
Ushkatta II, Spasskiy Most, Kent, Tagibay-Bulak, Dongal). 

This type of hearth corresponded to the early Indian special cult hearth 
ahavariiya (Mandel'shtam 1968: 126). Rectangular and round hearths have 
parallels in ancient Rome where the round hearth used for cooking was dedicated 
to the goddess of the domestic hearth Vesta; the square hearth was dedicated to 
male gods and the ancestors (Dumezil 1954). 

Type 4 is a two-chambered industrial hearth, consisting of a shallow pit to 
which a larger and deeper chamber is adjoined; sometimes a large vessel is 
placed there (Ushkatta II, Tursumbay, Kiimbay I, Sarafanovo) or we find a three- 
chambered hearth, sometimes covered and divided into three pits (Atasu, 
Shortandy-Bulak, Myrzhik, Entuziast, Upais, Ikpen', Tashik). 

Type 5 consists of shallow long slots extending along the longitudinal axis of 
the house, sometimes covered with stone and filled in with slag and ash 
(Bakhtinskoe, Kulevchi, Alekseevka, Tasty-Butak, Shandasha, Il'inka, Pokrovka 
III, Yavlenka, Atasu, Shortandy-Bulak, Tagibay-Bulak, Entuziast, Ikpen', Kent, 
Novonikol' skoe I, Petrovka II, IV, Yazevo, Kamyshnoe II). In Bakhtinskoe, 
Entuziast, Shandasha, and Atasu the slots or trenches were connected to the 
hearth and filled with coal, ceramics or metallurgical slag; they were sometimes 
covered by stone slabs; in Shandasha and Atasu openings for the nozzles of 
bellows led to a furnace, which indicates the industrial function of the hearth. 
They have analogues on Tazabagyab, Abashevo and Timber-grave settlements 
(Pryakhin 1976: 57, fig. 12; Gryaznov 1953; fig. 59, 60; Itina 1977: 202). 

Type 6 comprised an oven, probably with a dome made of clay bricks with 
an admixture of straw; they were rectangular, conical and round, and decorated 



46 CHAPTER FOUR 

with lines, circles and crosses (Sal'nikov 1951b: 131, fig. 17: 4). Such ovens 
served as a primitive furnace at Kipel', Shortandy-Bulak, Barmino, Yazevo, 
Kamyshnoe I, II, in Alekseevka, Yavlenka, Il'inka. Their analogues are known 
in the Timber-grave and Trzciniec cultures (Krivtsova-Grakova 1955: 76, 123; 
Berezanskaya 1974: 58-62). 

The number of hearths in a house ranged from one to eight. Wood, 
compressed dung and animal bones, which provided a high even temperature, 
were used as fuel judging from the finds of charcoal, burnt bones and dung. 

The interior of an Andronovo house was modest. At Tasty-Butak, 
Sadchikovo, Atasu, Kanay, and Kamyshnoe I there was found a two-meter wide 
bench along the walls forming a raised place for sleeping (Sorokin 1962a: 54; 
Potemkina 1985: 105; Maximova 1959: 93; Margulan 1959: 23). Analogues are 
known from Timber-grave and Tazabagyab settlements (Sinitsyn 1949: 199; 
Krivtsova-Grakova 1955: 75; Itina 1977: 200). In other houses there existed 
wooden plank beds (Alekseevka). Wooden tables are also known. In northern 
Kazakhstan there was sometimes a raised place built of clay near the hearths 
similar to a silon, the master's seat in a Pamir house (Andreev 1958: 461). 

In Arkaim, Tasty-Butak, Chaglinka, Petrovka II, Ikpen', Atasu, and Mirny III 
wells were found in the houses. They are pits about 1.5-2m in diameter, narrow- 
ing deeper to lm, and reaching the water table at a depth of 3-5m. The walls of 
the shaft were strengthened by wattling, layers of stones or by planks fastened by 
stakes (Zdanovich 1988: 183; Sorokin 1962a: 55; Orazbaev 1970; 1972: 154- 
162; Chemyakin 1978: 201; Margulan 1979: 174). The same wells, covered by 
sandstone slabs, have been found in the miners' quarters at Dzhazkazgan, 
Milykuduk, Sorkuduk, and Aynakol' (Margulan 1979: 268). The technology of 
making wells, learned by the Andronovans of Kazakhstan in the Petrovka stage, 
permitted the population of the Eurasian steppe to move off from the river 
valleys and master the deep waterless steppe, semi-deserts and deserts of Central 
Asia. 

The Andronovo house is comparable with houses of other cultures of the 
Eurasian Final Bronze Age over the vast territory of the steppe and forest steppe, 
from central Europe to western Siberia. The construction technique and planning 
decisions are universal over this whole zone. Those peculiarities that characterize 
individual archaeological cultures are seen only in terms of some secondary 
details. In the west, in central Europe we find post-built houses with a roof with 
two sloping surfaces, more rarely with four, erected as surface dwellings and 
smaller than Andronovo houses. These are known in the Unetice culture, in 
burials of the Tumulus culture, in the Fuzesabony-Ottomani culture, and they are 
preserved in the Hallstatt culture of the Early Iron Age, its variants being 
connected with different groups of Indo-Europeans: Celts, Illyrians, Thracians 
(Mongayt 1974: 52, 57, 63, 86, 87, 180, 192, 195). This house type is called by 
common Indo-European term: Slavonic dotm>, Sanskrit dam-, Avestan dsmana-, 
Latin domus. 

Along the Danube the Middle Bronze Age log houses combine the use of 
wattling with daub as in the Eurasian steppes (Kovacs 1977: fig. 6). The same 
peculiarity is traced in the Trzciniec culture in Poland (Kukharenko 1969: 49) 
and in the Ukraine (Berezanskaya 1972: 46-55; 1974: 22-42, 57-74: fig. 6-11, 
17-21). A semi-subterranian house with post-frame construction is typical of the 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 47 

Multi-roller Ware culture (Berezanskaya 1960: 27; Bratchenko 1976: 110-111). 
Similar also is the house type of the Abashevo culture from the Urals to the Don 
(Pryakhin 1973: 19-25; 1976: 11-24, fig. 2, 18) and Pre-Kazan' culture on the 
Volga (Khalokov 1969: 244, 257, 259). To the southeast of Andronovo territory 
are the similar houses of the Tazabagyab culture of the Aral region (Itina 1977: 
195-204, fig. 1,5-12, 30, 31). These derive from the migration of Timber-grave 
and Andronovo populations. (Peculiarities of the Tazabagyab house include its 
smaller area, hearth-pit with horseshoe-shaped side of pise and the semicircular 
floor before the entrance-a/raw). 

The closest analogues to Andronovo settlements and houses have been found 
in the Timber-grave culture of the Urals (Morozov 1989; Gorbunov 1989; 1992) 
and along the Volga (Sinitsyn 1949: 199, fig. 4; Krivtsova-Grakova 1955: 74-76; 
Merpert 1958: 104-118, fig. 13; Trubnikova 1953; 1958: 183-187, fig. 3) and 
along the Don (Gryaznov 1953; Pryakhin 1973: 45; Sharafutdinova 1967) and in 
the Ukraine where defensive walls have been discovered on the settlements of 
Kapitanovo, Il'ichevka and Usovo Ozero (A US 1985: 463-465). 

The most precisely reconstructed houses are known from the settlements at 
Suskan (Merpert 1958), Moechnoe Ozero (Trubnikova 1958) and Lyapichev 
Khutor (Gryaznov 1953). The area of the Timber-grave houses is somewhat 
smaller than that of the Andronovo. The houses are of timber-frame construction 
with a pitched roof supported by posts inserted along the center of the house in 
one, two (Suskan) or three (Moechnoe Ozero) rows; houses lacking evidence of 
posts are also known (Lyapichev Khutor). The entrances are corridor -like. The 
hearths comprise both open fires and large deep hearth pits, sometimes covered 
with stones; ovens have a brick dome. Along the Volga the houses were 
sometimes connected by passages. In the Ukraine, at Razdolnoe and Vovnigi 
there were stone socles formed from vertically inserted slabs along the walls of 
the house. Stone architecture began to emerge in the Final Bronze Age (AUS 
1985: 489-490; Leskov 1970; Sharafutdinova 1982). 

We can see then that the steppes and forest-steppes of Eurasia constitute a 
single zone according to settlement and house type of the cultures dating from 
the 17th to 9th centuries BC. Settlements are situated in similar topographic 
conditions along the banks of small rivers. The social function of such houses 
was to serve as a living place of a large, probably patriarchal, community with a 
common economy. 

Within the Eurasian zone Andronovo settlements and houses differ in some 
respect from other cultures. These differences include: 1) as a rule settlements 
exhibit regular planning and often have a large number of structures (as many as 
60-100); 2) the houses usually occupy a large area suggesting large family units 
and a high density of Andronovo population; 3) special ash-pits are found in the 
squares of settlements that suggest cults of fire and of the domestic hearth; 4) 
Andronovo houses often have walls covered with stone slabs; 5) a stone 
monumental multi-chambered house of regular rectangular form appears at the 
end of the Bronze Age in central Kazakhstan, which reveals the apex of 
architecture achieved by Eurasian populations in the Bronze Age; 6) 
Andronovans built special structures and annexes for industrial purposes; 7) the 
hearths of the Andronovo culture are numerous with respect to both type and 
function: they are divided into kitchen, ritual, and industrial, which reflects a 



48 CHAPTER FOUR 

higher level of craft development, especially metalworking, and a more 
developed domestic hearth cult; 8) special wells are sunk in Andronovo houses. 

The Andronovo house is then characterized by a stable combination of 
characteristics, uniform over the whole vast region of the Andronovo culture; in 
some aspects it differs distinctly from houses in other cultures. As domestic 
architecture is an important ethnographic trait, its homogeneity shows the 
cultural unity of the Andronovans themselves and serves as a significant 
criterium for defining Andronovo cultural unity. 

At the same time Andronovo unity according to settlement and house type 
should be viewed as a part of a larger Eurasian unity, stretching from central 
Europe to western Siberia. Settlements and houses of the various cultures of the 
Eurasian steppe and forest-steppe in the Bronze Age display uniform social fun- 
ctions, architecture, planning decisions and building techniques. It reflects in the 
first place a similarity of the economic level of development, and in the second 
place, a unity of house building traditions stemming from the Neolithic. Their 
origins are traced back to the early farming cultures of Europe, the Linearband- 
keramik, which was concentrated in Central Europe from Hungary to Germany. 
Settlements, protected from animals by bank and ditch, consisted of small semi- 
subterranean and large surface rectangular houses measuring up to 27 x 6m with 
post-construction and two sloping roof surfaces that are well reconstructed 
according to the evidence of excavation, e.g., Bylany. A model of a house with a 
pitched roof is known in Greece from Sesklo (Childe 1952: 103f). In the Tri- 
pol'e and related cultures rectangular surface houses with a framework covered 
with daub and having a pitched roof dominated, which is affirmed by house mo- 
dels from the settlements of Gumelnita (Denev) and Tripol'e (Kolomiyshchina) 
(Passek 1949: 79-106). But in the Tripol'e culture they also built semi-subter- 
ranian timber-framed houses and roofs seen at Luka Vrublivetskaya (Bibikov 
1953: 20-77). A long-term semi-subterranian house is known from settlements of 
the Sredniy Stog culture of the Dereivka type, where Dmitry Telegin excavated 
part of a house 10m long with an entrance-corridor and open hearth (Telegin 
1973: 39). The spread of subterranian dwellings with wooden constructions 
across the steppe is indicated by the character of the burial chambers of the Pit- 
grave culture which included roughly hewn logs (Gorodtsov 1905: 199, 302). 

An essentially different house type is known from the Kelteminar culture of 
Central Asia and the Ust'-Narym in eastern Kazakhstan: oval in shape, surface, 
short-term post -built cabins (Tolstov 1948: 61; Gulyamov et al. 1966: 26-28, fig. 
8,9; Vinogradov 1981: 148-150, fig. 59; Chernikov 1970: 12-14; Chalaya 1971: 
13). In Kazakhstan, in the Neolithic-Eneolithic period, there was another type of 
house — a small rectangular or round semi-subterranian dwelling, with a conical 
roof made from poles and skins (Borubas, Damsy, Botai, Kozhai (Margulan 
1959: 13-14; Zaybert 1993: 21; Kalieva 1998: 225)). This type of house is 
characteristic of hunter-gather-fisher societies. Only in the late Neolithic 
settlement of Vishnevka in north Kazakhstan does one encounter a house which 
is similar to the Andronovo house type involving a semi-subterranian area, 18 x 
7m, with log walls, supporting posts, pitched roof over the center and a stone- 
covered hearth (Zaybert 1973: 106, fig. 39). 

The settlement and house type of the steppe and forest-steppe zone of Eastern 
Europe and the Urals approaches that of the Central European type but differs 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 49 

fundamentally from the houses of Near Eastern farming communities. Our 
evidence of the Near Eastern house type derives from its Neolithic and 
Eneolithic cultures. From the 7th and 6th millennia BC we find a type of house 
with walls built of pise and rarely of stone; the flat roof dominated everywhere. 
As the house evolved, it showed a tendency toward increasing the number of 
rooms, the enclosure of the building around a central square, and residences 
forming a continuous block building divided by narrow streets. Such houses, 
consisting of a main room 9-12 square meters, and some storage or domestic 
annexes 2-3 square meters, with a flat roof, and pise -built domed ovens are 
known already from Jarmo, Hassuna and Yarym-Tepe in Mesopotamia 
(Braidwood 1952; Lloyd and Safar 1945; Oates 1976; Childe 1952: 170, 177; 
Afanas'eva and D'yakonov 1970; Masson 1976b; Ivanova 1981; Rapoport 1966; 
Kuz'mina 1988a). From the Halaf period (5th millennium BC) there appeared 
round tower-like tholos structures, 4-7m in diameter, with a dome-shaped roof. 
From the 4th millennium BC two-storied houses and monumental buildings 
constructed from sun-dried brick and mortar appeared in Mesopotamia. Multi- 
room houses built of clay with a flat roof are also known from the Neolithic and 
Eneolithic periods in Anatolia (Catal Hoyiik) (Mellaart 1967), Syria-Palestine 
(Jericho, Ras-Shamra) (Kenyon 1960), Iran (Sialk 1, 2, Shahr-i-Sokhta) (Ghirsh- 
man 1938; Tosi 1969) and in northern India in the settlements of the pre-Harap- 
pan and Harappan cultures (Sankalia 1974). Single-room surface houses, pise- 
built, that accommodated a small family appeared in the Jeitun period in the 
south of Central Asia (Berdyev 1969; Masson 1971: 103-104; 1976b: 117-119) 
and in the Eneolithic period at Kara-depe and Geoksyur these were replaced by 
tholoi and multi-room houses with a flat roof and inner yards (Sarianidi 1962; 
Khlopin 1964: 67-87; Masson 1962: 157-175; 1976b: 119-120) that were used 
for a residential unit of several small families. This type of house survives 
unchanged to the present time (Amir'yants 1981). These houses and type of 
settlements with continuous quarters of buildings are characteristic of central and 
western Asia, Iran and India, and are fundamentally different from the houses of 
the pastoral tribes of central Eurasia. 

Difference between the two traditions is observable in all cultural traits. 
There is no Near Eastern influence on the traditions of house building in the 
Eurasian steppes during the 3rd to 1st millennia BC. This observation is signif- 
icant when considering the hypothesis advanced by V. V. Ivanov and T. V. Gam- 
krelidze about the mass migration of the ancestors of Indo-Europeans from the 
Near East through the Central Asian deserts into Eastern Europe. The evidence 
of domestic architecture does not support this hypothesis nor those of J. Nichols 
(1997) and V. Sarianidi (1998). 

We can see then that Eurasia exhibited two house types for the Eneolithic and 
Bronze Age: an Indian-Near -Eastern (pise- or brick-built, multi-room house with 
small rooms, flat roof and a built oven) and a Central Eurasian (wooden post- 
built single-room dwelling, sometimes with clay covered walls, pitched roof and 
open hearth). 

Ecological conditions preconditioned both main house types: the abundant 
precipitation in Europe and the presence of forests led to the creation of a timber- 
built house with a pitched roof; in Asia the lack of precipitation and the absence 
of forests allowed people to build clay houses with a flat roof. House size and 



50 CHAPTER FOUR 

interior plan came from social and economic conditions. The large patriarchal 
family survived for a long time in the steppes and led to a common household, 
requiring a large single -room house, but in the Near East there were small 
families utilizing small rooms. Once the house type had been developed in the 
Neolithic, these building traditions remained as features of various cultures and 
were carried into new regions over the course of migrations. 

In the Near East and in India the multi-room house, built of pise or brick, and 
with a flat roof dominates from the Eneolithic onwards. Here the appearance of 
an alien house type with pitched roof is associated only with the Indo-European 
peoples: in cult constructions of Phrygia and Lydia (Akurgal 1961) and in the 
architecture of the Greek polises of Asia Minor. It is an important marker of the 
alien character of the Phrygians, Lydians and Greeks in the Near East and 
indicates their migration from Europe. Within the ethnographic record, the type 
of house with a pitched roof and timber-post construction is known among 
different Iranian-speaking peoples of Iran and Afghanistan, and among some 
Indic-speaking groups of India. 

A direct continuation of the building traditions of the Timber-grave/ 
Andronovo cultures can be seen in the culture of the Iranian nomads such as the 
Saka, Sarmatians, and Scythians, the likely descendents of the Bronze Age 
pastoral tribes in the Eurasian steppes. Despite a shift to nomadic pastoralism, 
monumental architecture continued in the steppes in the early Iron Age. Accor- 
ding to the evidence from Kamenskoe, Bel'skoe, Pasterskoe, Matroninskoe, 
Nemirovskoe, Tarasovskoe and other settlements, the Scythians surrounded their 
settlements with defensive walls, constructed large semi-subterranian houses 
with stepped corridor-/iaM(iw.s and surface houses of ground-post-built 
construction whose ridge-pole rested on posts along the longitudinal axis. 
Miniature house models with a pitched roof have been recovered from a burial in 
the kurgan at Zhurovka. As in the Timber-grave and Andronovo houses, the 
walls were constructed from longitudinally laid logs or poles fixed by vertical 
posts (rarely from posts inserted near each other); wattle and daub is also rare. 
They seldom used stone in construction, e.g., on the Bug. Hearths were made as 
before in the form of fire pits and ovens from clay bricks. A large ash-pit might 
be found in the square of the settlement (Grakov 1971: 61, 62, 123, 124, 136, 
138, 141, 146, 147, 152, 153, figs, on pages 62, 63, 122, 137, 155, 158). 

The settlements of the Asian steppes are poorly known and our knowledge 
of house construction derives largely from the evidence of structures in kurgans. 
In Arzhan there was discovered a magnificent construction in the form of a 
round platform, 80m in diameter and 3m high, built around the central 
framework of the royal grave. It consisted of 70 timbers and was covered with 
flat logs and surfaced with stone slabs (Gryaznov 1980: 10, 14, 15, fig. 4-7). 
There was a mound, 1 10m in diameter, which was surrounded by a stone kerb of 
upright slabs. 

At Pazyryk the royal burials were placed in timber-frame constructions, 
covered by flat logs supported by post uprights; the roof was covered by bark 
and branches. In kurgan 5 a kibitka of felt (the upper covering of a wagon) was 
discovered, placed on a vehicle (Rudenko 1963: 78-80, tables 2-9). In Besshatur- 
skiy kurgans 1 and 6 the chambers were made of horizontally laid logs, fixed in 
place by vertically inserted posts; the logs were of Tian-Shan fir. There was a 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 5 1 

door and narrow entrance corridor floored by logs with reed mats and a large 
piece of felt on top; the mound with its encircling stone ring was 52m in dia- 
meter and 8-9m high (Akishev and Kushaev 1963: 31-40, 49, fig. 10-26, 33-48). 
Just like the Andronovans, the Saka also used stone widely. In central Kazakh- 
stan they made graves covered with stone slabs (Margulan et al. 1966: 311). The 
same construction methods were employed by the Sarmatians. In their graves 
there was a timber frame or a layer of wooden planks and then on top were 
placed large pieces of felt, reed, bast and birch bark. In the Pokrovskiy kurgan 
horizontal logs of a timber frame were held by timber uprights. In Leninsk and 
Alebastrovaya an oak roof was supported by upright posts. In Usatovo, Boaro, 
and Pyatimary, graves contained rectangular timber frames of oak or birch. In 
Preobrazhenka and Tarabutak round timber constructions were discovered. 
These structures "imitated the rectangular surface dwellings of the Timber-grave 
houses and round tents" (Smirnov 1964: 85-89). Sarmatians as well as Saka used 
stone such as round kerbs and slabs for roofing graves (Smirnov 1964: 85, 89, 
90). Both the domestic architecture and grave -type of the Andronovans survived 
among the Pamir Saka. In Tegirman-say, Kyzyl-Rabat and Vorukh they buried 
the dead within stone circles that enclosed stone cists, which were set in a pit and 
covered by stone slabs (Litvinsky 1972: 134-135). This rite is preserved in some 
districts of Afghanistan, in Ishkashim, Vahan, Darvaz and in Yagnob among 
Iranian peoples of the mountainous regions of Tadzhikistan (Andreev 1927: 53; 
Andreevand Polovtsev 1911: 17-18; Rakhimov 1956: 69). 

A house form that is very close to the Scythian prototype is preserved among 
the descendants of the Scythians — the Ossetes of the Caucasus. The Ossetic 
house differs considerably from the saklya of the neighboring Caucasian peoples. 
In South Ossetia there is an archaic type of house — a large rectangular semi-sub- 
terranian nykkamd divided into two parts, a cattle shed and a living place with a 
hearth in the center (Kaloev 1971: 153). E. A. Kaloev rightly compares Ossetic 
semi-subterranian houses with those of the Scythians. Research indicates that the 
formation of this house type is several thousand years old and can be connected 
with Timber -grave and Andronovo house -building traditions. Surface houses are 
more wide-spread in Ossetia. Wood, stone, and rarely clay, are used as const- 
ruction materials. The houses are timber-framed with walls made of vertical logs 
and wattle and daub. The houses often consist of a single large rectangular room 
(x«jar), built for a large patriarchal family. The floor of the xse^ar is earthen; 
there is an open hearth coated with stones which is a sacred place for the family. 
The hearth divides the house into two: the right part for the men and the left for 
the women. A small dark larder {k'sebits) adjoins the women's section. The 
Ossetic house roof is pitched, supported by walls and posts; the main post is the 
astsewykkag sagyn^; it stands in the center of the house and has cult significance. 
The roof consists of a main ridge pole that extends over the long axis and that is 
supported by the central post of the ridge beam (astsewykkag axarag). Rafters 
rest on the beam and then brushwood, straw, especially reeds, are placed on 
them; everything is often covered with earth or coated with clay from above. 
There is a smoke hole (erdo) in the roof (Kaloev 1971: 145-163, fig. on pages 
148, 151, 159). Sometimes a pyramidal-stepped dome is constructed of the chor- 
hona type; a house with such a roof is called an erdojali-saxli (Pchelina 1930: 
19; Il'ina 1946: 1 1, fig. 2). This type of roof has been borrowed by the Georgians 



52 CHAPTER FOUR 

and Armenians from the Ossetes along with the Iranian construction terms 
(Mamatnazarov and Yakubov 1985: 199). In some cases the entrance is long, 
with an overhead cover supported by posts (Kaloev 1971, fig. on page 151). 
Plank beds are made along the walls, which are covered with skins and carpets. 
Horns of wild goats and rams and sacrificed horse skulls are placed at the 
entrance. Sometimes separate rooms for newly-weds (wat) are built as annexes 
to the x«jar. V. I. Abaev traces the etymology of this word from early Iranian 
(Kaloev 1971: 152). Wats are connected with the xse^ar along the same axis, 
resulting in a long rectangular house (dargxse^ar), analogous in plan to the long 
Andronovo house with annexes. Near to or separate from the xa;jar may be a 
guest-room (wcizcegdon 'friends' room'). Each house belongs to a large 
patriarchal family. In a settlement, sometimes in the whole valley, the kinsmen 
share a single cemetery and sacred place (jivar). In an Ossetic settlement the 
houses are not built close to one another as in Caucasian auls. They are 
dispersed, each house having its own wattle-fenced yard {ksert) with timber- 
framed stores. Scholars have noted that the Ossetic xa;jar is quite similar to the 
large house described in tales of the Narts (Abaev 1949: 55; 1965: 130; Kaloev 
1971: 150). Comparison with houses of the Timber-grave and Andronovo 
cultures indicates that the origin of this type of house dates back to the middle of 
the 2nd millennium BC. The xajjar is analogous to the Timber-grave/ Andronovo 
house in terms of its social function (place for a large family); in plan (large 
single chamber rectangular house, sometimes with annexes along the same axis); 
in construction technique (timber-frame ground post-built construction, pitched 
or pyramidal-stepped roof); in hearth type; and in interior layout (hearth in center 
and plank beds along the walls). All this leaves no doubt that there is a direct 
genetic connection between the modern Ossetic house and the dwellings of the 
Timber-grave and Andronovo cultures. 

Andronovo architectural traditions developed on the eastern end of the 
Eurasian steppes among the Iranian peoples of Tadzhikistan. Their houses differ 
greatly from the classical Near Eastern dwellings. Tadzhiks in the mountainous 
regions of Tadzhikistan and in Afghanistan build single-room rectangular 
timber-frame houses (hona), and often make semi-subterranean houses with 
wooden or stone walls. Sometimes the house consists of two chambers: a deeper 
winter chamber and a higher-ceiling summer chamber. Pise beds (sufd) are 
arranged along the walls (Andreev 1927; 1958; Voronina 1959; Pisarchik 1970; 
Logasheva 1981a: 124-134). 

Houses of the Pamir type in Afghanistan, Badakhshan, Nuristan, north-west 
India and districts of Pamir (Fig. 9: 6, 8-10) are of special interest (Huf, Vahan, 
Shugnan, Rushan, Vanch, Yazgulem, Ishkashim, Karategin and Darvaz) 
(Kislyakov 1936: 131; 1939: 166; Voronina 1951a: 256-271; Andreev 1958: 
267-273, 420-480, fig. 22, 95-97, 102, 103,111-114, 116-122; Pisarchik 1975: 
159-168; Mamatnazarov and Yakubov 1985: 183-202; Logasheva 1981a: 132- 
134). These regions were settled by peoples speaking relict Iranian languages 
and archaic dialects such as Vahan, Shugnan, Rushan, etc. (Gryunberg and 
Steblin-Kamensky 1974: 277-279) and the Indo-Iranian population of Chitral in 
Pakistan where such a house type was described by Aurel Stein (1912: fig. 20). 
The Pamir house is a large rectangular building (chod), at Huf from 35 to 75 
square meters in area, and at Vanch and Vahan up to 120 square meters. A 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 53 

peculiarity of the Pamir house is its distinctive roof (chor-hona). A square 
wooden frame is placed above the center of the house on four supporting posts, 
and a succession of ever smaller timber frames overlap one another to the top. As 
the frames are laid on each other diagonally they form a stepped vault with a 
light-smoke opening in the center; the opening is called a day-ruz and it is closed 
at night (Andreev 1958: fig. 22, 57, 58). As mentioned above, this type of roof 
was already known in the Timber-grave and Andronovo house. It is preserved in 
the ritual architecture of Kazakhstan (Basenov 1959: fig. 13,19) and in the dom- 
estic architecture of Central Asia it has survived till the present time (Voronina 
1951b; Pisarchik 1954: 273; 1975: 159-168; Zhilina 1982: 161, fig. 5). In Afgha- 
nistan and eastern Turkestan this construction technique was translated from 
wooden architecture into stone: the Buddhist cave temples of Bamiyan exhibit 
skeuomorphic ceilings cut into rock in their sacred places in imitation of the 
chor-hona roof type (Godard and Hackin 1928; Hackin and Carl 1933; Puga- 
chenkova 1963: 69-70) as also in the cave temples of Kizil in Turfan (Le Coq 
1925). 

The Pamir house is semi-subterranean with an earthen floor, earthen beds and 
a stone socle, made of stones laid without mortar; in Rushan the stones are set on 
edge (Andreev 1958: 432-434). Then posts are erected, beginning with the main 
post, the king post or ha-setan, on the main beds, then central posts supporting 
the overlapping layers of the chor-hona, and other posts along the axis of the 
house, in the corners and along the walls supporting the roof beams. The timber- 
frame chor-hona is constructed above the center of the poga; the remaining roof 
is constructed from beams and poles (behm) which are first covered with grass 
and reeds and then covered with earth. A closed corridor (dalidz) leads into the 
house, the floor level is higher there than the poga floor in the house. There is a 
summer room iyuzhrd) and a guest room (kush-hona) near the dalidz. Beds (noh) 
rise above the floor; they are divided by special short stone or wooden partitions 
into special functional sections. At the entrance there is varguht, a room for 
lambs, gau-hona, room for a cow; along the long house wall there are beds with 
an open hearth where women cook, and adjoining beds (chirezak) with a quern 
and vessels containing food; opposite the hearth there is the ceremonial part of 
the house with the main post and beds (sar-tek-noh), where guests are met and 
the head of the family lives, and adjoining beds were in the past the place of the 
bridegroom. A two-tiered hearth is open, the main pit is coated with stones and 
clay; it is oval and it rests on bedding. There is a shallow pit for the ashes on the 
floor in front of it (this house type is also known from Andronovo houses). 

The hearth is the domestic center of ritual where sacrifices are made. A clay 
platform with a stone near it serves as the master's seat. In the center of the floor 
there is a sump (ob-hin) into which liquids are poured. The interior of the house 
is decorated with skins and carpets on plank beds. Mountain goat horns hang at 
the post; sometimes the main post of the house is covered with plant ornament 
which symbolizes the tree of life. 

The Pamir house is constructed by a group of relatives and the construction is 
accompanied by sacrifices of a ram, rarely a bull, and feasts at the laying out of 
the foundations of the house, the erection of the walls, the raising of the main 
post (ha-setan) under which the leg of a sacrificed lamb is placed, at the raising 
of the roof beam over which the blood of a sacrificed ram was poured. Finally, 



54 CHAPTER FOUR 

when the family moved into the new house, blood was poured on the threshold, 
races were held, there were shooting matches (sar-asnd) and the slaying of a 
goat (Andreev 1958: 432, 437, 443, 445). Sacrifices also accompanied the 
construction of an Andronovo house: bulls and ram skulls and legs have been 
recovered inserted at the entrance or under a post and the walls on many 
settlements. A large patriarchal family of 40 to 80 people inhabit a Pamir house 
(Kislyakov 1936; Andreev 1958: 431). 

Pamir villages contrast with those of the Near East by having their houses 
dispersed as among the Andronovans. Sometimes temporary houses were built 
near the main house for summer use: this included a pitched-roofed annex 
{chapkand), a pitched tent chodar, a raised platform of branches {kappa), a 
fenced enclosure (dikori) and open hearths. Stone -built sheds for cattle are made 
on summer sites. Traces of such constructions are also attributed to the Timber- 
grave and Andronovo settlements. 

Pamir and Ossetic settlements and houses present, in the first place, a more 
complete picture of the life of the Andronovans, the planning of the interior of 
the house, the hearth and the ritual role of the supporting post, etc. In addition, 
the similarity between these modern houses and those of the Timber-grave and 
Andronovo cultures attests the genetic continuity of house-building traditions. 
The domestic architecture of other of the smaller Iranian-speaking peoples 
differs considerably from the classic Near Eastern type which demonstrates the 
alien character of their house. The winter house of the Kurds, Iranian pastoralists 
in Asia, is a one-room rectangular subterranian dwelling with stone walls and 
timber uprights along the longitudinal axis; the posts support a stepped roof 
constructed from square timber frames with an opening for light and smoke in 
the center. The surface of the roof is covered with branches, earth, and clay. On 
the earthen floor in the middle of the house is an open hearth pit lined with 
stones. Sometimes storehouses are built as annexes to the house (Aristova 1981: 
78-79). In the Gilyan and Mazanderan provinces of Iran, people have sub- 
terranian and surface ground-post timber houses with a pitched roof covered with 
reeds and bark (Logasheva 1981b: 106-107). Wooden architecture dominates in 
Nuristan as well. As in an Andronovo house the walls are made of horizontally 
laid logs held in place by timber uprights and a roof supported by timber posts 
(Vavilov and Bukinich 1929: 119; Robertson 1906: 487, 489; Logasheva 1981b: 
134-136). The Iranianized Jemshids of Afghanistan have pise house walls, but in 
terms of planning and the pitched roof they preserve early house building 
traditions that are not of Southwest Asian origin (Gafferberg 1948: 131). 

Thus, in the farthest reaches of Iran and Afghanistan, Iranian-speaking 
pastoral nomads preserve planning decisions and construction techniques that are 
alien to Near Eastern traditions of domestic architecture. They have no local 
sources in the Neolithic or Bronze Age of Iran and Afghanistan but they do have 
analogues in the Eurasian steppes, which prove the northern genesis of their 
creators. 

The origin of north-central Indian building traditions is probably the same. 
Here we find a single-storey rectangular structure, sometimes single -roomed 
with a timber frame and a simple or double-sectioned roof; the ridge beam is 
supported by posts along the longitudinal axis (Fig. 9: 7). The roof is covered 
with straw, reeds and other roofing materials (Guseva 1981: 153-158). There is a 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 55 

large room for men and guests in the house {baithak, chkho pal or chaupal), 
sometimes this 'friends' room' is a separate building. In multi-room houses of 
the rajput there is a large common room or pol. Rooms for individual small 
families and cattle sheds are sometimes annexed to the house under the same 
roof. Cattle are sometimes kept in a shed {vara). Such houses are for a large 
patriarchal family. They are dispersed and isolated from the houses of the lowest 
castes. Such houses are especially typical of the north-west and central states: 
Gujarat, Rajastan, Uttar-Pradesh, Madhya-Pradesh, Bihar and Maharashtra. In 
other districts of India quite different domestic architecture obtains that is similar 
to the Near Eastern pise or brick house with a flat clay roof. Ground-post-built 
houses and pitched roofs have no analogues in the Neolithic and Bronze Age 
architecture of India. Houses in the Dekkan and along the Ganges exhibit round 
and rectangular huts with a conical or pyramidal roof; they are found in Tekka- 
lakot, Navasa, Navdatoli, Chandoli, Inamgaon, Saipai and Nagda (Sankaliya 
1974: passim, Shchetenko 1968: 32, 137; 1979: fig. 11, 37, 38). In northwest 
India there were houses of the Indo-Near Eastern type already in the pre- 
Harappan period (Kot Diji); in the Harappan towns, however, multi-chamber 
brick houses with flat roofs made up regularly planned quarters of continuous 
buildings (Makkay 1951: 40-51, tables 1-4). Building techniques and planning 
decisions worked out by Harappan architects have survived in India to the 
present in the house of Indo-Near Eastern type. As the house type of northern 
and central Indian coexists everywhere with the Near Eastern type house and 
because it is built in different geographical zones, its specific character is 
probably preconditioned not by ecological factors but by the domestic architec- 
tural traditions alien to the creators of the Harappan culture and brought from the 
outside. The isolated nature of the northern and central Indian house type in 
South and Southwest Asia and the presence of its analogues and prototypes in 
the pastoral cultures of the Eurasian steppes lead us to conclude that such a 
house type was brought to India by groups of Aryas who migrated from their 
homeland. The distribution of houses belonging to the central Eurasian type 
within India suggests two waves of Aryan migrations into India from the north- 
west: 1) through the north-west regions of the country, and 2) along the Gangetic 
plain. An important proof of this hypothesis is the fact that the houses of north 
and central Indian type are connected with quite specific ethnic and social 
circles. They are built by representatives of the higher castes, sects and ethnic 
groups within which the institute of a large patriarchal family is preserved: 
rajputs, gujars, kankan brahmans, djats, etc. The majority of scholars accept that 
these groups are ethnically and genetically connected with the arrival of the 
Vedic Aryas in India or with later waves of Central Asian tribes related to the 
Aryas. Variants of the northern and central Indian house type are also character- 
istic of the pastoralist Ahirs who migrated from the north, of the Meo, the ethnic 
and caste group who arrived from Central Asia in the Kushan period and the 
Pashtuns of Afghanistan (Guseva 1981: 159). All this connects the appearance of 
the ground-post-built house with pitched roof with the house building traditions 
of the Eurasian steppe pastoralists. 

Thus, an analysis of the Andronovo house helps to establish its Central 
Eurasian origins, its genesis in the steppe zone is traced from the Eneolithic, and 
the tradition is further reflected in the cultures of the Saka, Scythians and 



56 CHAPTER FOUR 

Sarmatians as well as among Indo-Iranian peoples of the present. There is no 
influence of the Near Eastern architectural tradition, neither in construction 
technique nor in planning, in the steppe zone of the Bronze Age. Consequently, 
the evidence of the Andronovo house does not suggest a migration of the 
Andronovans from the Near East nor does it contradict their attribution to the 
Indo-Iranians. The latter is supported by some linguistic data. The word for 
'house' in the Indo-Iranian languages is the common Indo-European term: 
Sanskrit dam-, Avestan dsmana-, Slavonic dom-b, Sanskrit dam-, Latin domus, 
Greek domos (Burrow 1976: 98; Elizarenkova and Toporov 1995: 512). The 
word for 'house' in other Indo-European languages is applied to the house type 
of central Eurasia with its pitched roof from which it follows that the house of 
the ancestors of the Indo-Iranians also belonged to this type. The word for 'door' 
is common Indo-European: early Indian dvdr- 'door' and synonymous with 
'house', Avestan dvar- 'gate, door', Latin fores, etc. The word 'fence' is also 
common Indo-European: Sanskrit grha-, Avestan gsrsda-, Slavonic grad-b, 
Lithuanian gardas, English yard (Gamkrelidze and Ivanov 1984: 744). With 
respect to locating the Indo-Iranian homeland, it is significant that this word was 
borrowed from Indo-Iranian into the Finno-Ugrian languages. 

There is another common Indo-European term in the Indo-Iranian languages 
to denote 'fence, stone wall' (Abaev 1979 III: 114). An important Indo-European 
word used to denote 'cattle shed' is Sanskrit stharta-, Russian start from the 
Indo-European stem *sta- 'stand'. Russian start denotes not only 'nomad camp' 
but also 'residence'. The name of the Persian capital, Persepolis, originates from 
parsa-starta- 'nomad camp of the Persians'. Later the word starta- was borrowed 
into the Turkic languages of Kazakhstan, Turkestan, etc. (Abaev 1979 III: 153; 
Elizarenkova and Toporov 1995: 512). 

The word skambhd- is of common Indo-European heritage (Renou 1939: 49). 
T. Ya. Elizarenkova and V. N. Toporov (1995: 512) showed that in the Rigveda 
it meant 'support, pillar, column', in the Iranian languages — 'supporting beam'. 
In the Rigveda (4.13.5; 9.74.2; 9.86.46, etc.) we find the metaphor of a 'sky 
column'. In the Atharvaveda (10.7.8), skambhd- is deified as a creature who 
supports the universe, sky and sun — the equivalent of the 'world tree'. 

The Indo-Aryan (and not Iranian) form, in the meaning 'pillar' and with the 
mythologized associations with a sacred 'column', was borrowed into the Finno- 
Ugrian languages. The image of the magic Sampo mill, an analogue of the world 
tree in Finnish mythology, originates from it (Erdodi 1932: 214-219). A common 
Iranian name for 'house' originates from the verb kan- 'to dig', i.e., it reflects a 
semi-subterranean type of house (Abaev 1956; Herzfeld 1941; Oransky 1976: 
163). Benveniste suggested that the early common Iranian verb vi-da- 'to build a 
house' has the prefix vi and the root da- 'to create' (Benveniste 1955: 301; 1958: 
65). This term is found in the Avesta. Its ancient meaning is preserved in 
Shugnan wiSurt (from Old Iranian *vidana- 'ceiling, overhead cover'). The chor- 
horta roof of the Pamir house, which originates in the Andronovo culture, is 
denoted by this word. The chor-hona construction is a most important labor- 
consuming stage in the building of a house (Andreev 1958: 441). 

V. A. Livshits analyzed the data concerning the house as described in the 
Avesta (1963: 141, 505; 1998: 271). This house had a post construction with a 
roof supported on posts; their Avestan names are mit- and sturta- and these have 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 57 

correspondences in Vedic (Elizarenkova and Toporov 1995: 512). Derived from 
stiina- are the Persian and Tadzhik sutun and Shugnan set an, which is applied to 
columns and pillars, to the supporting posts of the Pamir house in particular 
(Andreev 1958: 445). The house was illuminated through a light opening in the 
roof {raucana-, cf Tadzhik ruzan, Shungan raz). This term is used in Pamir 
houses for the light-and-smoke opening in the chor-hona roof (Andreev 1958: 
437). In addition to timber, clay (istya-, cf. Tadzhik hisht) was used in house 
construction. In modern Central Asian architecture this term also denotes 'brick' 
(Pisarchik 1954: 223), but in ancient times it was applied only to clay. In Pamir 
and Ossetic houses clay is used in wall plastering and in the roof covering 
(Andreev 1958: 434, 440, 441; Kaloev 1971: 153). The interior of the Avestan 
house included the gatu- 'bed'; according to V. A. Livshits (1963: 141) this was 
a clay sufa. Thus, the house of the ancestors of the Iranians was a large semi- 
subterranian ground-post-built construction that supported a roof with a light 
opening through it, and which contained a hearth and sufa. 

In the Avesta, in the ArdvTsur Yast (5.101), dedicated to the goddess Anahita, 
it is said: "At every stream there is a solid built house, it is light, with a hundred 
of light openings, well-made, with a thousand pillars, firm, with ten thousands of 
supporting pillars." From all the various house types that are archaeologically 
attested in Eurasia during the 2nd millennium BC this description of Anahita' s 
house is closest to the large Timber-grave/ Andronovo house. A reconstruction of 
a large house from the settlement of Atasu, suggested by A. Kh. Margulan 
(1959: table I), illustrates this description. 

The evidence for the Vedic house is also of great interest. In the Atharvaveda 
(3.12.1) there is a spell dedicated to a protector-deity of the house Vastospati: 

1 . Just here I fix a firm house 

Let it stand in peace, sprinkling fat! 

Let us enter you, oh house, 

With healthy men, with beautiful men, with unharmed men! 

2. Just stand here firmly, oh house, 

Rich in horses, rich in cows, rich in joys... 

9.3 

11. He who fixed you, oh house, 

Gathered forest trees, 

He made you for descendants... 

13. ...Homage to cows, horses, 
To the one born in the house! 

14. You cover fire inside, 
People together with cattle. 

In the text (9.3) props, supports, overhead covers, a sacrificial place, a fire 
hearth, a place for women are mentioned (9.7). T. Ya. Elizarenkova translates the 
term sala- as 'hut', but the text alludes to a timber construction built for men and 
cattle; that is why I accept L. Renou's translation 'dwelling, house'. Renou 
(1939) analyzed the early Indian texts connected with house building. His 
research was continued by R. Pandey (1965), H. Bodewitz (1977/78; 1979) and 



58 CHAPTER FOUR 

W. Rau (1983). The Indo-Aryans chose a good place before building a house, 
drew on the ground a Purusa, the legendary first man who was sacrificed. Then 
they measured a rectangle from the Purusa, marking with a furrow the borders of 
the future house oriented according to the quarters of the world. The house 
length is 12-16 paces, its width is 10-12 paces (which would yield a general area 
of approximately 100 sq meters which could hardly be described as a hut). Then 
they dug a foundation ditch, pressing the excavated soil over the edges. The next 
action was to dig knee -deep pits in the floor for the wooden support posts 
(sthuna- or skambha-) that were erected vertically in the center or along the 
perimeter. The text speaks of 'many pits', 'a large number of pillars'. Then they 
put the overlapping crossbeams on the supporting posts at right angles to each 
other, and finally they set the rafters of the pitched roof resting on the stupa- or 
'roof ridge' - this term later came to denote a ritual construction (Renou 1939: 
187, 198). W. Rau (1983) suggested that the stupa- originally referred to the 
supporting pillar, but according to the context this is not correct. In the 
Brahmanas it is said that the builders set the overlap of the two sides of the 
pitched roof of the house on a central beam-ridge (Rau 1983). Four central 
supporting posts that support the rectangular overlapping log frame, as in the 
Pamiri house, are mentioned; moreover, one of the pillars is called 'king's post'. 
A light opening, bed, seat, and mats are also mentioned. Wood is the only 
building material; neither brick, pakhsa, nor stone are mentioned in the texts. 
Stones are used only for the construction of the altar and hearth. Twigs, brush- 
wood, reed, mats (only in later texts straw) are used as roofing material. The 
door (dvdr-) turns on two supports in front of the door. There are two types of 
open hearths in the house: a round domestic hearth for the women and a rect- 
angular cultic hearth for the men. There are light household structures, a kitchen, 
a shed for the chariots near the house (Renou 1939: 491; Rau 1983: 20-59). 
Some terms denote cattle shed or sty {urva-, gotra-) (Elizarenkova and Toporov 
1995: 512). Thus, according to both texts and the lexicon, when the Vedic Aryas 
arrived in India they constructed large rectangular semi-subterranean houses of 
post -built construction that were analogous to those of the Andronovo culture. 

The rituals accompanying house construction were also similar. The building 
of an Andronovo house began with a solemn ritual: two to three vessels of milk 
were offered, more often a bull or ram was sacrificed. Their meat was eaten by 
the participants at the ceremony, the head and legs with skin symbolizing the 
dead animal and the vessels were ritually deposited in important places of the 
future house — under the corner support post, under the entrance threshold or at 
the hearth. Until recently the Iranian Mountain Tadzhiks in the Pamirs and the 
Ossetes in the Caucasus had the same rite. 

There are also examples of human sacrifice, often of small children, buried 
under the floor or in a special annex, that are known from the Andronovo cul- 
ture. This rite can be explained in terms of the early Indians' belief that the uni- 
verse was composed of the parts of the body of the first man (Purusa). The whole 
house is then a microcosm, a model of the universe in miniature. It means that 
the construction of a house requires the same type of sacrifice. The Vedic Aryans 
in India rejected this bloody rite. They simply drew a Purusa on the place of the 
future house. But the rite of ritual killing was preserved for the construction of 
temples and royal palaces in some parts of South Asia to the 20th century. 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 59 

The most important and most difficult moment in construction was the 
raising of the roof. In the Atharvaveda (3.12.6) there was a special spell 
associated with the act: 

'Climb the pillar correctly, rafter!' 

In modern India a feast is held with the raising of the roof ridge (pashita) as it is 
also done in Ossetia and in the Pamirs where the beam is sprinkled with the 
blood of a sacrificed ram. The supporting 'king's post' is also important. It is 
symbolic of the axis mimdi that, according to the early Indo-Iranians, connects 
the world of men with the sky, a tree of life providing prosperity for the family. 
Both in India and Ossetia this support post was deified; the house god Sasryzajd 
was associated with him (Abaev 1979 III: 91). In memory of old beliefs this 
pillar was often covered with rich plant ornament in Ossetia, the Pamirs, and the 
Hindukush; on spring holiday of the New Year Nouruz, it is decorated with dog 
rose flowers. The patriarch usually sits at the pillar; on wedding days the bride 
and bridegroom are seated there. 

Moving into a house was also accompanied by special rituals common to 
different Indo-Iranian peoples. The housewife was the first to enter the house: 
she brought vessels filled to the brim, then the hearth was solemnly lit from a fire 
taken from the previous house, and spells were pronounced: 

'Homage to the moving fire, 
Homage to your Purusa' 
(A iharvaveda 9.3.12) 

The domestic hearth was a sacred place for the family: the dead were carried 
round it before the funeral; a bride was taken round it when bringing her into the 
house; a solemn oath was said above it; and a threat to ruin the hearth is still the 
most terrible in Ossetia. The master and mistress of the house made sacrifices at 
the hearth in honor of the gods and ancestors. At the feast of Fravardin in 
remembrance of the dead, the Avestan people sacrificed to the ancestors and 
invited them to the feast. 

In the Avesta (Yasna 62.3-7; Videvdat 18.18-22) it is said that the domestic 
hearth and ashes are sacred, and every morning it is prescribed to take ashes with 
special ceremonies to a particular place (Dhalla 1922: 66, 67, 155). In India the 
phrase 'to set a fire' denotes the separation of a son from the family of his father 
to build his own house. In the 'Laws of Manu' the throwing of the ashes of the 
hearth into the wind is proscribed and they must be gathered into a pile. This 
Indo-Iranian tradition explains the construction of special ash piles found within 
Andronovo and Scythian settlements. 

In general, a comparison of the archaeological and linguistic data about the 
monumental house supports the ascription of the Andronovans to the Indo- 
Iranians. 

The word 'house' in the Indo-Iranian and other Indo-European languages 
denotes not only dwelling but also the family and its property. According to the 
Avesta a large patriarchal Iranian extended family, as with a Greek oikia, lived in 
the house. The patriarch was at the head of the family (like the Roman pater 



60 CHAPTER FOUR 

familias, Vedic dampati-, Avestan nmano.pati- 'master of house' and his wife 
nmano.paOni- 'mistress of the house'). They undertook religious functions and 
made sacrifices. A group of close relatives on the father's side constituted the 
nafa-, an agnatic group whose patronymy united up to a hundred adult males and 
was connected with common pasture and land and a custom of mutual help. Not 
less than 15 houses/families made up a vis-, a clan and the settlement of the clan 
ruled by an assembly headed by an elected vispati- (D'yakonov M. 1961: 60-61; 
Livshits 1963: 140, 143, 504; Gafurov 1972: 54-55). The name vispati- comes 
from Indo-Iranian pati- 'husband, master' (< Indo-European *potis); it is found 
in Sanskrit, Avestan and Lithuanian (Burrow 1976: 25, 98), and the word vis- 
has correspondences in Slavonic vbsb, a village with its areas of economic 
significance. Survivals of this ancient organization have survived among the 
Mountain Tadzhiks and Ossetes. In the Pamirs a large family of 40-80 members 
might live in one house. In Vanch a family of 55 members lived in a house 121 
square meters in area (Kislyakov 1936: 755-788; Andreev 1958: 431). Related 
families occupied one settlement or valley, helped each other and participated in 
common rituals. People speaking one language occupied one or several neigh- 
boring mountain valleys. Such a structure is well represented in the ethnography 
of the Ossetes: a large patriarchal family lives in one house, uniting from 20-40 
to 100 people headed by a xse^aru-xicaw 'head of the house' and his wife or 
sefsin 'mistress of the house'. Kinsmen sharing a common sacred place and 
cemetery and the same surname are members of one patronymy or kin group and 
live in a settlement. Sometimes the whole valley belongs to one kin group. 
Kinsmen own land together, they are interconnected by a custom of mutual 
assistance and they constitute an assembly (Kaloev 1971: 133-136, 191-193, 
203-205; Magometov 1962: 968). 

The next unit in the Avesta is the unification of the clans into a zantu- 'tribe', 
a name which also has Indo-European correspondences, e.g., Latin genus. 

If we take such ethnographic data into consideration when examining the 
distribution pattern of Andronovo sites, we may suggest that the larger groups of 
Andronovo sites constitute local variants, which can correspond to the ancient 
tribe that occupied a territory and was separated from other tribal territories by a 
largely unoccupied zone. Densely populated micro-districts situated along small 
rivers and included within a local variant belong to individual clans; the 
settlement unites a patronymy that shares a common cemetery and common 
sacred place (a large settlement can belong to the whole clan). The individual 
house provided the habitation for a large patriarchal family, numbering 30-50 or 
more people. The ascription of such a number of family members is 
independently confirmed by comparison of the number of burials in a cemetery 
and the number of houses in a settlement where the average life span of the 
Andronovo people is estimated at 33 years (Kuz'mina 1974c). 



House Type II 

Type II houses are light timber-frame constructions, rectangular, polygonal or 
rounded in plan (Fig. 10). The number of such types of construction increases in 
late Andronovo settlements. In central Kazakhstan at Atasu there were 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 61 

workshops and houses of oval or round form, surface or slightly subterranean, 
with light timber-frame walls (Margulan et al. 1966: 208; Kadyrbaev 1983: 137; 
Kadyrbaev and Kurmankulov 1992: 25, 44, fig. 19). Analogous houses are found 
in Myrzhik; in Buguly I house 22 was round (Margulan 1979: 187; 1992: 58). 
Oval houses were found at Ikpen' (Tkachev 1991: 10). At Suukbulak the walls 
of wooden house 1 are of irregular rectangular form. A. Kh. Margulan recon- 
structs it as polygonal (Fig. 10: 5; Margulan et al. 1966: 248, fig. 126). Surface 
houses of oval and rectangular form and measuring 50-100 square meters in area 
have been discovered at Novonikol'skoe I and Petrovka II in north Kazakhstan 
(Zdanovich 1975a: 9, 10; 1988: 36, 37, 44, 45; Zdanovich S. 1979: 9; Malyutina 
1991: 142). In the Urals, at Yazevo there is an oval house, 10.8 x 4.2 square 
meters and 0.2-0. 3m deep (Potemkina 1985: 38). At Dangal in central Kazakh- 
stan a round house, 70 square meters in area, is surrounded by slabs inserted into 
the ground (AO 1983; 1985: 507). In Semirech'e at Bien, round houses, 31-70 
square meters in area, are set into the ground and surrounded by a ring of stones 
on a clay foundation (Karabaspakova 1987: 95, 96). The best studied houses are 
those from the late Andronovo settlement at Chaglinka (Fig. 10: 7, 8; Orazbaev 
1970: 134, 136-139). House 9 is a round semi-subterranean structure with 
timber-frame construction; the walls consist of horizontally laid blocks and ver- 
tical posts on which the timber frame of a conical roof rested. House 13 is two- 
roomed and slightly subterranean: one chamber has timber-frame construction, 
the other is oval with uprights forming a twin wall with wattling; the light 
conical roof was supported by walls and by raised posts from which pits remain 
(probably a square frame was placed on the posts and used as smoke opening). 
House 14 was also two-chambered with rounded corners (perhaps octagonal) and 
timber frame construction. On the settlements of Kipel', Alekseevka, 
Zamaraevo, etc., post pits and narrow slots were preserved as also traces of the 
light surface constructions of wattled walls or pailing. 

Several slightly sunken round houses without posts and with a sand bank 
around the floor and a hearth in the center were discovered on the late 
Andronovo site of Dzhanbas 34 in Khorezm (Itina 1977: 105-106). Light timber- 
frame houses were also constructed in the western steppe. On the eastern 
Trzciniec settlement of Pustynka there were surface houses consisting of 
vertically erected posts and wattling (Berezanskaya 1974, 43, 51, tab. X, XI: fig. 
15). Post pits and clay daub with wattle imprints from light surface constructions 
are also known from settlements of the Pre-Kazan' culture (Kalinin and 
Khalikov 1954: 198). At the late Timber-grave settlement at Seragoz there were 
two surface yurt-like round houses, 8m in diameter with stone foundation walls 
and pise -built hearths (Boldin 1981: 232). Thus in the late Bronze Age across a 
wide zone of steppe and forest-steppe there was spread a type of light timber- 
frame construction, vertically divided in two, many sided or round in plan, either 
surface or slightly sunken, and with a light conical roof. 

The origin of this type of house is connected with the evolution of the large 
rectangular semi-subterranean house; a number of structures (e.g., at Chaglinka) 
indicate transitional forms. The invention of a light house in the late Bronze Age 
was the major innovation in the culture of the Eurasian steppes and provided for 
the transition to pastoral nomadism. 



62 CHAPTER FOUR 

The light timber-frame Andronovo houses of the late Bronze Age can be 
compared to the tents of the modern Iranian-speaking nomads of Iran, 
Afghanistan and the Near East. The most primitive type of house has remained 
among the Afghan-Durrani and partially among the Toimen and Charaimak 
tribes. This is atent or gidzhi (Logasheva 1981a: 142-143; Ferdinand 1969: 131- 
135). Several vertical poles are erected over a rectangular base, 6m x 3-5m in 
area. The central poles are higher than the side ones; they are covered with felt or 
cloth, resulting in a conical roof; the walls are tightened over, the floor is 
covered with felt. The similar but larger gedan is spread among the Baluchi in 
Iranian Baluchistan and Iranian and Afghanistan Sistan. The poles (achag) are 
erected in three rows, five in each row, over an area of 10m x 3.5m; the central 
pole is the highest. Felt sections are drawn over the poles, forming a conical 
upper part. The sections making up the side walls are fastened to its sides. An 
entrance is covered with cloth (Gafferberg 1964; Ferdinand 1964: 197; 1969: 
129). The same principle is characteristic of the house of the Iranian Kurds, who 
migrated through the Near East, and part of the Charaimaks, who are territorially 
quite distant from them and move about in Afghanistan. Their house also 
consists of vertically driven poles, arranged in rows, but differing from the gedan 
because all the poles of the central row are higher than the side ones, so that the 
roof becomes pitched (Aristova 1965; 1981, 78-82; Vilchevsky 1958: 195-196; 
Feilberg 1944: 85, 179; Ferdinand 1969: 131-133; Logasheva 1981a: 143). 

Another variant of the timber-frame house is found partially among the 
Kurds, Baluchi, migrating in Baluchistan and in south Afghanistan, and among 
Iranianized Jemshids and Khazaris in Afghanistan. This is the kappa house or in 
its more developed form, the chappari. In comparison with previous houses they 
are round in plan. Their timber-frame consists of poles driven into the ground, 
the tops of which are bent and bound together; in the chappari they are fastened 
to an upper wooden hoop that carries the conical vault which is covered with a 
large piece of felt. The lower part of the chappari walls are covered by reed 
wattling (Gafferberg 1948: 131-140; Logasheva 1981a, 1981b: 118, 139, 143; 
Barth 1965; Feilberg 1944). 

Some maintain that the Iranian tents derive from the Bedouin type (Aristova 
1981: 81). It is difficult to agree with this conclusion. The tent among the 
Semitic-speaking nomads of the Near Eastern Bedouins is a one -part dwelling. A 
large rectangular section is fastened to four poles, a space with flat roof is 
constructed which is trapezoidal in section; one part of the tent remains open, the 
opposite part is covered with a curtain. Sometimes the tent is stretched across 
two supporting props in the form of an irregular trihedron (Amir'yants 1981: 71- 
72; Fielberg 1944). The types of houses of Iranian-speaking nomads in Iran and 
Afghanistan are basically different in construction from the Bedouin tents of the 
Near East. Bedouin tents are on two or four poles, Iranian tents utilize many 
poles driven in rows in a rectangle or circle. In structure Bedouin tents are 
without a separate roof section and are single-chambered; Iranian tents are 
double-chambered, they have vertical walls and a separate roof. With respect to 
roof form, Bedouin tents have either a flat top or a single slope; Iranian tents are 
pitched, pyramidal or conical. In terms of roofing method, Bedouin tents employ 
one panel in making up the roof and two walls; Iranian tents have one panel for 
the roof and side walls made of separate large pieces of felt, mats, etc.; all of this 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 63 

speaks for a quite different genesis for the black tents of the Semitic-speaking 
Bedouins and the houses of the Iranian nomads. In spite of considerable 
differences in form, Iranian tents have basically a similar post construction, and 
are vertically divided into two sections. Their prototypes are the light Bronze 
Age houses of the Eurasian steppes. Both are two-chambered, consist of vertical 
walls and sloping roof, have timber uprights as their foundation, walls of 
wattling or other light materials, a roof supported by wall supports and a central 
supporting post, sometimes set on a rectangular open frame and simultaneously 
serving as both a light and smoke opening (the burnt frame of such a house was 
found on the Andronovo settlements of Bishkul' IV and Pavlovka). The three- 
row arrangement of supporting posts of the gedan of the Baluchi and especially 
in the tent with a pitched roof among the Kurds and Charaimaks repeats the plan 
of the Timber-grave/ Andronovo house of the Moechnoe Ozero type (Trubnikova 
1958, fig. 3) and its lighter surface variant at Pustynka. Only the character of the 
roof differs as it is made of felt and not of wood. The round Iranian tents or 
chappari can probably be connected with the late Andronovo round surface 
houses. 

The similarity in construction of the light surface houses of the Eurasian 
steppe belt in the late Bronze Age with the tents of Iranian-speaking pastoral 
nomads of Iran and Afghanistan suggests that these nomads came to the south 
from the Eurasian steppes, bringing with them the type of house that was 
established in the south Russian steppes beginning with the Eneolithic. This 
similarity suggests the probable time of the Iranian migration at somewhere 
between the 13 th and 10th centuries BC. 

This conclusion is confirmed by linguistic data: the early Proto-Iranians 
meant by vi-da- 'to build a dwelling of the Andronovo subterranian type'; 
however, in the Middle and Modern Iranian languages vi-da- only denotes a 
'tent, awning, marquee' (the name of the tent, qidan, among the Baluchi nomads 
originates from it, cf. Benveniste 1955: 301; 1958: 65; Kuz'mina and Livshits 
1987). It follows that the house building traditions of both settled and nomadic 
Iranian-speaking peoples originate from a common source and they can be traced 
back to the Bronze Age Andronovo culture. 

Light surface houses, household and ritual structures were known to the 
Vedic Aryans as well (Elizarenkova and Toporov 1995: 491). In the Atharva- 
veda (9.3) and in the later Jaiminiya-Brahmana (1.70-72), in the Grhyasutras and 
Srautasutras, short-term collapsible constructions are described {Atharvaveda, 
commentaries by T. Ya. Elizarenkova 1976: 383; Renou 1939: 484-491; Pandey 
1965; Bodewitz 1977/78; 1979: 77-82; Rau 1983). They are modest in size (7- 
9m x 4.5m x 7m), they are of rectangular or round form and they are made of 
poles tied at the top and mats bound with ropes. While transported, they are 
untied and carried in a vehicle. Terms for their construction, derived from the 
root ma- 'measure', 'build', are not quite clear; 'support', 'upper part', 'cover 
fastening', 'rafter', 'binding ropes' and 'mats' are all mentioned. 

In an Atharvaveda spell (9.3.4ff.) we read: "...We unbind everything at the 
thatching and around the fastenings... we unbind everything bound... we unbind 
your loops... straw covered hut... you stand on earth... I unbind everything 
bound to your mat." "Oh Hut, we shall carry you as a bride wherever we wish". 



64 CHAPTER FOUR 

These texts refer us once more to the question of the origin of the yurt. A. M. 
Orazbaev (1970: 142-146) made a detailed comparison of the polygonal timber- 
frame and round house construction on a Chaglinka settlement with the con- 
struction of a modern Kazakh house (shoshala) and with a nomadic yurt, and he 
concluded that the "prototype of the modern yurt ... could be the house construc- 
tion of tribes living in the late Bronze Age". This conclusion was affirmed by K. 
A. Akishev, M. K. Kadyrbaev (1977: 222) and by S. Zh. Narynov (1980: 170- 
173). The Timber-grave/ Andronovo traditions of house building continued to 
develop among the Iranian Scythians, Sarmatians and Saka. The Saka construc- 
ted lighter buildings as well. The invention of felt was an important contribution 
of the Andronovo period. Its usage was known in Pazyryk and Besshatyr. It is 
important that in Pazyryk felt was used for vehicle covers, tent walls and floor 
coverings (Rudenko 1953: 78-80); in Besshatyr the roof was covered with a 
large piece of felt (Akishev and Kushaev 1963: 58). Timber- frame construction 
is preserved among the Sarmatians (moreover, in Preobrazhenka and Tarabutak 
round structures were found, cf. Smirnov 1964: 85-89). According to K. F. 
Smirnov these burial constructions "imitated the rectangular surface timber- 
frame houses and round tents". The Scythians had two types of light houses: 
rectangular and round in plan. Walls for both house types consisted of vertically 
erected poles or posts across which was set a wattle and daub fence, a mat or a 
large piece of felt. There was sometimes a support post in the center of a round 
tent, which supported the light conical roof. Kurgan #1 at Kostromskaya 
provides a brilliant illustration of such a rectangular structure. Four main posts 
were erected at the corners. Vertically placed logs provided a timber-frame on 
which was set horizontally lain beams. Leaning against this framework were 
posts which met at the center to form the outline of a pyramidal roof (Artamonov 
1966: table 5). 

Another type is seen in a light temporary dwelling represented in a painting 
in the burial chamber at Anfesterios (Fig. 10: 3). This is a house of pyramidal 
form with two high raised central posts (Rostovtsev 1914: 170-173, table ci). 
According to S. I. Vaynshtein (1976: fig. 6, 3,4) it can be reconstructed as a 
truncated pyramid; in the upper third of it there is a border in the form of a 
rectangular wooden frame; above this is a rectangular frame connected to the 
poles of the roof part which simultaneously served as a light and smoke opening. 
L. G. Nechaeva (1975: 11-13) justly considers this frame to be a prototype of the 
shangrak of a yurt whose genesis can be related to the light-smoke openings of 
houses with a pyramidal-stepped vault of the Timber-grave and Andronovo 
cultures. L. G. Nechaeva (1975: 13) did not examine the Bronze Age evidence, 
but she studied in detail Scythian houses and came to the conclusion that they 
can roughly be called primitive yurts. A. M. Khazanov (1975: 271) stated that 
the Scythians "definitely knew the yurt", but S. I. Vaynshtein (1976: 43) argued 
that the Scythian tent cannot be called a yurt because it probably did not have 
folding lattices which make up a characteristic feature of the modern yurt. 

The yurt is the primary and most perfect house type of the Eurasian steppes 
nomads (Fig. 10.9). It is used by all Turkic-speaking and Mongol peoples, as 
well as among the nomads of Afghanistan — Iranian-speaking Tadzhiks, 
Firuzkukhs, north Taimens and Iranianized Jemshids and Khazari (Kharuzin 
1896; Gafferberg 1948; 1953; 1964; Vilchevsky 1958; Aristova 1965; 1981; 



SETTLEMENTS AND DOMESTIC ARCHITECTURE 65 

Logasheva 1981a; 1981b; Ferdinand 1960; 1964; 1969; Barth 1965). The name 
for yurt in Iranian is alachik or chador, Khazari khanai khyrga, Turkish oi, gara- 
oi. Its walls are made of separate folding lattices (in Turkish tarem, kanai), 
plaited from twigs, reed or strips of wood {sagana), bound with straps. A vertical 
round wall (kerege) is made of several lattices. The roof is made of poles or slats 
(uk), their sharpened ends ikalam) are set in a wooden hoop {shangrak) which is 
fixed from above. At the sides they attach the kerege to the forks of the poles in 
such a way that there appears a cylindrical vaulted framework. A shangrak, 3- 
5m in diameter, is used as a light-smoke hole. The diameter of a yurt is 9-1 5m. 
The kerege walls are covered with pieces of felt, the vault is covered separately, 
the floor is covered with felt and carpets, and the door is curtained or made of 
wood (Margulan 1964: 1-12). An advantage of the yurt, which consists of 
several folding lattices, is that it is convenient to transport and easy to install. 

The time and place of the origin of the yurt and its prototypes have not been 
settled yet. The first researcher of nomad dwellings N. N. Kharuzin (1896: 46) 
believed that the yurt had a long evolution, its present form being established 
only in 17th century, and that the Siberian chum was its prototype. S. I. Vayn- 
shtein (1976: 51) concluded that the prototype of the yurt was the hemispherical 
cabin (shalash) of the Huns, with a frame made of twisted willow twigs 
(Vaynshtein 1976: 45, fig. 6, 7). B. Kh. Karmysheva (1956: 23) believed that the 
proto-type of the yurt was the same construction as that of the Karluks. This 
point of view has been widespread. Many scholars believe that the yurt devel- 
oped in Central Asia among Turkic tribes. It was then borrowed from them by 
the Mongols (Kyzlasov 1975: 173) and Iranian-speaking nomads of Iran and 
Afghanistan (Logasheva 1981b: 112; 1981a: 138-139). At the same time it is 
thought that many-sided houses originated among the Turks not long ago when 
they made the transition to settled life (Vaynshtein 1976: 50-51). The opposite 
opinion has been expressed by investigators of the nomads of the Eurasian 
steppes (Orazbaev 1970; Purveev 1982; Kuz'mina and Livshits 1987). 

To settle the question of the origin of the yurt one should compare the house 
types of the modern nomads of Asia, construct a typology and compare them 
with the yurt. The Siberian chum is a single-section dwelling, its framework 
formed by poles conically converging at the top (Fig. 10.2). The cabin or shalash 
of the Huns is also vertically a single -section dwelling. Its framework is formed 
by arches which converge hemispherically at the top, or it is woven from twigs 
in the same hemispheric manner. The 'black tents' of the Bedouins are also a 
vertically single-section dwelling consisting of two or four poles covered by 
rectangular panels to form an irregular trihedron or truncated pyramid 
(Amir'yants 1981: 71-72, fig. 10.1). Iranian nomads have quite different houses. 
They are bi -partite and consist of vertical walls and a separate roof. 

The Iranian tent, including that of the Scythians, is closest to the construc- 
tion of the yurt and can be viewed as its prototype. Both the yurt and the Scyth- 
ian tent are vertically bi-partite dwellings with vertical walls and separate roof. 
In the Scythian tent one can find the most important parts of the yurt: 1) frame- 
work wall construction consisting of a base of vertical poles; 2) walls filled with 
wattling or mats; 3) a truncated-conical or pyramidal roof element supported on 
the upper edges of the wall and by a frame above; 4) a framework construction 
used as a light and smoke opening which provides a point of fastening for the 



66 CHAPTER FOUR 

roof; 5) presence of a door; 6) felt covering; and 7) felt and pile carpets in the 
interior. 

The only invention left to the creators of the yurt was to transform the 
unfolded wattle sections to folded lattices or tarem. This innovation could be 
introduced only by a people acquainted with the Scythian tent. Its construction 
developed in the south Russian steppes from the Eneolithic; in the early Iron Age 
together with the Scythian triad (horse-riding, weapons, and animal art style) it 
probably reached Central Asia and was adopted by local Turkic tribes. The 
earliest dated image of a yurt -like dwelling with conical roof is a painting from 
Great Boyarskaya dating to the 2nd century BC (Devlet 1976: 9, tables 5-7). But 
the painting is too schematic to draw conclusions about the presence of folding 
lattices in this dwelling. The earliest known yurt image with folding lattices and 
edges is a Zoroastrian ossuary of the 4th-5th centuries AD in the Samarkand 
museum (Vaynshtein 1976: 49, fig. 3, d, e). This type is quite analogous to the 
yurt khanai-khyrga with conical top found among the modern Iranian-speaking 
Khazari of Afghanistan (Gafferberg 1953). 

Early Indian texts, the Atharvaveda (9.3) in particular, indicate that the Vedic 
Aryans were already well-acquainted with the type of folding transportable 
dwelling constructed from poles and mats, close to the Scythian and modern 
Iranian nomad tent. 

These data prove that it was Indo-Iranian peoples who created dwellings of 
the proto-yurt type. 



House Type III 

The third type of Andronovo house comprises kibitkas, large covered vehicles on 
solid wheels. According to Hertel (1925) such covered vehicles represented the 
sky palaces of the gods, vimana (see section "Transport"), described in Vedic 
literature. This vehicle was the main house -type of the Aryas in the Vedic period 
(Rau 1977, 1983; Elizarenkova and Toporov 1995: 488, 490, 511; Elizarenkova 
2000). 

Thus, the evidence of the history of the Andronovo house confirms its 
identification as Indo-Iranian and supports the opinion of linguists who argue 
that the Indo-Iranians migrated to India and Iran from the Central Asian steppes. 



CHAPTER FIVE 
CERAMICS 

A large body of ethnographic research in various parts of the world has shown 
that the technology of ceramic production differs according to ethnic group 
where the traditions of domestic production are transferred to and regularly 
preserved within related groups (Franchet 1911; Cardew 1969). For this reason 
ceramics are considered to be one of the most diagnostic ethnic indicators that 
permit one to establish the genetic relationship between population groups, even 
after distant migrations and large chronological breaks between complexes; cer- 
amics also help to reveal the process of assimilation of different ethnic groups. 

The analysis of ceramic production is critical because ceramics constitute the 
bulk of archaeological material and provides the classificatory basis of Androno- 
vo sites; the technology of pottery manufacture defines to a certain degree the 
form of a vessel and serves as an important indicator of the ethnic group. 

V. A. Gorodtsov (1922) and M. V. Voevodsky (1930; 1936) demonstrated 
that the pottery of the Neolithic and Bronze Age East European cultures was pro- 
duced by the coil technique by building up strips of clay coils, 4-6. 5cm long, or 
by constructing spiral coils or subsequent courses of entire rings, or employing 
existing pots as a model. This was evident from the analysis of Andronovo and 
Kayrak-Kum ceramics (Krivtsova-Grakova 1948: 142-143, fig. 29; Sal'nikov 
1951b: 132, 133; Litvinsky 1962: 234, 235). M. P. Gryaznov established that the 
Begazy Dandybay vessels of central Kazakhstan were produced by a different 
technique of beating out the form from a single block of clay. E. F. Teplovod- 
skaya also studied the ceramic technology of central Kazakhstan (Kuznetsova 
and Teplovodskaya 1994). M. N. Komarova (1962) identified the main types of 
ornamentation on Andronovo vessels and traced their distribution over various 
zones. S. V. Zotova (1964, 1965) established two different principles of creating 
Andronovo decoration over either a straight or oblique grid (Fig. 12). 

I have analyzed the vessels of the whole Andronovo period according to a 
single system. Acquaintance with the different methods of ceramic production 
based on ethnographic trips in highland Tadzhikistan (1951-1953), the ceramic 
centers of Gidzhuvan and Rishtan in 1958, and study of pottery-making skills in 
a Samarkand workshop under the supervision of D. Dzhurakulov helped to 
inform my analysis. The basis of my classification of Andronovo ceramics was 
closed burial complexes. A single program of analyses was utilized for the exam- 
ination of every vessel. These comprised: 

1. The composition of the clay and the character of any admixture in the clay; 

2. The technique of vessel formation with special attention to the base and the 
order of coil placement; 

3. The form of the vessel; 



68 CHAPTER FIVE 

4. The surface treatment which included: a) the principle of the construction 
of the decoration; b) the main elements of ornament; c) the placement of the 
elements according to zones and their combinations; and d) the technique of 
placing the ornament (Table 1, Fig. 2). 

After the vessels were divided into groups on the basis of similar 
characteristics the combinations found within closed burial complexes was 
examined. Finally, those complexes characterized by ceramics with a stable 
combination of similar traits were grouped into types. Diagnostically important 
criteria for singling out the types proved to be the technology of production and 
the form of the pot-like vessels, the nature of the tempering, whether the 
ornament was drawn over an oblique or straight grid, and the technique of 
manufacture. Undecorated pots were not informative. 

As a result of the analyses four pure types were singled out: Petrovka, 
Alakul', Fedorovo, and Alekseevka and a large group of mixed types. 

Petrovka 

Sites of the Petrovka type are traditionally dated to the 17th 16th centuries BC, 
and are situated in the west of Andronovo territory in the Urals, Tobol, the 
steppe and forest-steppe of western, northern and central Kazakhstan as well as 
the settlement of Tugai near Samarkand (Avanesova 1996; Fig. 69: 7-16). 
Petrovka pots are made from quite plastic clay with artificially added tempers of 
shell, sand, gravel, chamotte, talc and sometimes mica to permit the escape of 
moisture during firing (Smirnov and Kuz'mina 1977: 13-4, 17, 26; Zdanovich G. 
and S. 1980: 187). The ceramics from the Urals are characterized by an 
admixture of talc. The color of Petrovka pots ranges from ash-gray to brown and 
black. Vessel walls are 8-13mm thick. On the inner surface of some of the 
vessels there are the traces of the application of a fabric (woolen threads 8- 
1.6mm thick or horsehair). Another vessel served as the mold and was covered 
with the fabric and the base was attached to the walls from the outside. After 
drying, the mold was removed and the upper portion of the pot was placed on top 
from the inside: this resulted in a sharp edge or rib that formed where the two 
elements joined to form a biconical vessel. 

The relation of the height of the upper part of the vessel to the lower varies 
from 1: 4 to 1: 6; in northern Kazakhstan vessels are squatter (1: 2-3) and have a 
biconical bell-shape with a flanged rib. Both their form and ornamentation emp- 
loying applied rollers and bosses are similar to the vessels of the late Catacomb 
and Multi-roller Ware cultures (Smirnov and Kuz'mina 1977: 29-32, fig. 9). 

For repelling moisture the outward surface of a vessel was sometimes given a 
clay slip which resulted in the slurred ornament on some vessels. 

The ancient potter understood the peculiarities of constructing a two-zoned 
vessel and emphasized them through ornamentation: there was a separating band 
on the shoulder, a semantically important ornament was placed on the body 
while a different motif decorated the upper zone of the vessel. Petrovka ornam- 
ent was made on a vertical grid, mainly by a combed stamp with large wide 
teeth; in northern Kazakhstan sometimes the comb was used in a rocker motion. 
In a few cases cord, caterpillar stamp, shell, or small spade-like indentations 
were used which reflect the preservation of Neolithic traditions. A large percen- 



CERAMICS 69 

tage of Petrovka vessels were covered by fluting, horizontal wavy grooves that 
stood as an independent ornamental device. The predominant ornamental pat- 
terns were horizontal and vertical herring-bones, zigzags, and isosceles triangles; 
from time to time there was a broken-band rhombus and herring-bone divided by 
vertical lines over the body with a lattice or straight swastika on the bottom. 

The preference for vertical and horizontal herring-bone patterns, often spread 
over the entire body, reflects the Neolithic tradition. It must be stressed that the 
transition from a Neolithic bowl to the biconical pots of the Early Bronze Age 
occurred similarly among the Poltavka, Petrovka and Potapovka tribes by 
placing an upper section on the vessel (Smirnov and Kuz'mina 1977: 34, fig. 6; 
compare Smirnov 1959: fig. 12.6, 28.15; Sinitsyn 1958: fig. 45). In some comp- 
lexes there are Abashevo pots. Vessels analogous to Petrovka including Poltav- 
ka, Abashevo, and Late Catacomb, are found in Bashkiria at Al'mukhametovo 
and in burials of the Potapovka type on the Volga at Potapovka, Alekseevka 2, 
Utevka 6 and on the Don at Vlasovka and Kondrashkinskiy (AD 1977: 159, 198; 
1980: 133; Vasil'ev et al. 1992: fig. 2.2-10, 4.19, 5.1-4, 6.12; Pryakhin 1992: fig. 
1.16; Vinnikov and Sinyuk 1990: fig. 28. 1,2, 29.1-4). These are similar with 
respect to the technologies of production, the usage of shell and sand temper, the 
imprinting of decoration in zones by a plain and toothed stamp, and the motifs of 
ornament. It reflects the genetic closeness of the Potapovka and Petrovka po- 
pulations, formed simultaneously in the forest steppe from the Don to the Trans 
Urals with the participation of the Poltavka, Abashevo and Late Catacomb popu- 
lations in their ethnogenesis (Smirnov and Kuz'mina 1977; Bochkarev 1991). 

A later group of Petrovka sites (Raskatikha, Kenes, Aksayman, Satan, Nurtai, 
Krasnaya Krucha) is defined by the western orientation of its burials; the gradual 
diappearance of the Multi-roller Ceramics and Abashevo cultures, and vessels 
with an additional section forming a tripartite vase. This form is characteristic of 
early Alakul' ceramics. 

AlakuV 

Sites of the Alakul' type, dated to the 15th— 13th centuries BC, are situated in the 
steppe and forest-steppe zone in the west of Andronovo territory in the same 
areas as Petrovka (Figs. 11-13; 15, 16). 

Unlike Petrovka pots, the fabric of Alakul' pots does not contain shell 
temper. Gravel, large grain sand, quartz, and mica were used as tempers. In the 
Urals Alakul' vessels contained talc as did the Petrovka pottery from this region. 

The walls of Alakul' vessels measure 7-8mm, and are thinner than in 
Petrovka. Many Alakul' vessels preserve distinct fingerprints on the inner sur- 
face, especially under the rim and near the base. These suggest that the vessels 
were made by women. Traces of textile impressions that covered the mold are 
often seen on the inner surface of the vessels, especially in the Urals. The fabric 
was placed over the firm surface of a ceramic mold, which means that Alakul' 
vessels were produced in the same way as Petrovka ones. Some pots were 
formed differently: the potter began with the base and built up the lower part of 
the body by applying coils (bottom start, according to A. A. Bobrinsky 1978). 
Sand that has stuck to the clay is sometimes seen on the base of the vessel. Ap- 
parently, the rest of the vessel was completed on a sand-covered stand. Flat stone 



70 CHAPTER FIVE 

discs have been found in settlements which could have served as such stands. 
This is supported by ethnographic parallels: in the mountains of Tadzhikistan 
craftswomen use such a convex form of round stone stands (tova) or clay or 
pressed dung {binniki or benuk) when a pot is constantly turned on the stand 
(Voevodsky 1930: 67; Grigor'ev 1931: 62; Peshchereva 1959: 31, 36, 45, 46, fig. 
3-5, 8, 9). In north-western Pakistan they use a wooden ring (Rye and Evans 
1976, tables 1,3). 

The next coil which began the inward slant was attached to a mold from 
above as was done in the manufacture of Petrovka vessels, and a ridge, a peculia- 
rity of Alakul' ceramics, was also made at the place of attachment. A rim coil 
with an inward slant was attached to this portion from above, and an edge 
appeared at the place of attachment. 

Variants A and B are defined according to the relation of the maximum rim 
diameter to the body diameter; subtypes are defined on the basis of the height of 
the upper part of the vessel. Alakul' pots have a standard form and proportions. 
Due to the additional rim section the Alakul' pot has considerably different pro- 
portions to the Petrovka pot: the relation of the upper part above the edge to the 
body ' s height is usually 1 : 2, the diameter of a rim is equal to that of the body. 

After forming the surface of the pot it was smoothed with a cow rib: with one 
hand a craftswoman turned the vessel on the stand, with another she held a tool 
for smoothing the surface. Finds of such polishers are numerous on Andronovo 
sites. Their usage as tools for making pottery is confirmed by ethnographic 
parallels from Central Asia and Pakistan (Voevodsky 1930: 68; 1936: 62; 
Peshchereva 1959: 30-32; Rye and Evans 1976, table 20). In Tadzhikistan such 
bone or wooden tools are called lisik (from Iranian lis- 'lick, to make smooth'). 

Finally, the outer surface was consolidated and sealed against moisture by 
burnishing. Andronovo sites have yielded flat pebbles that were used as 
polishers. 

Ornament was placed on the vessel when its surface was damp. The construc- 
tional peculiarities of a tripartite vessel were understood and stressed in the orna- 
mentation of the pot: the ornament in the three zones was separated by division 
lines and each zone had its own specific form of ornamentation. In western 
Alakul' variants the neck zone was devoid of ornament while the semantically 
important decoration was placed on the upper part of the body. 

The ornament was impressed by a medium- or large-toothed stamp, rarely 
with a square or rectangular cutting edge. Incized ornament is rarely met on 
Alakul' vessels. On many vessels two different stamps coexisted: the main elem- 
ents of ornament were made by a combed-tooth stamp and the divisions were 
indicated by a stamp with triangular or oval teeth. Fluting was used to form 
independent elements but they differ from the Petrovka ceramics as they only 
separated the rim and especially the neck zones on Alakul' ceramics. A woman 
first drew the separating lines between the zones and then she filled in the 
ornament over each zone, from top to bottom, then from left to right. There was 
no preliminary setting out of the ornament within the zones, and, consequently, 
the rapport is broken on many vessels. 

Alakul' decoration, like that found on Petrovka, was set out on a straight grid 
(Figs. 2; 12; 16). Hatching of ornamental bands is horizontal, and the main 
elements of ornament were the same as in Petrovka ceramics: zigzag, vertical 



CERAMICS 71 

herring-bone, isosceles triangles with their apexes up, and rhombuses. There are 
also specific Petrovka motifs, e.g., broken bands descending from the rib (Cher- 
nyaki); herring-bone divided by vertical lines; and modeled bosses on the body. 
Basal ornaments are preserved on some Alakul' vessels. Typically Alakul' 
motifs include a pyramid, opposing large and small triangles, a triangle with a 
M-shaped figure at the apex, and especially a simple meander and its variations. 
All but the pyramid elements are characteristic of the early Timber-grave ornam- 
ental complex which indicates the related nature of both ornamental traditions 
and the similar direction of their development. Alakul' decoration differs from 
Timber-grave only with respect to its three zone division and the large variety 
and complexity of the combintation of elements. 

The final operation in pot production was firing. Andronovo ceramics from 
sites in Turkmenia were fired at temperatures ranging from 500/600 to 800 
degrees; in Chust — 750/800 degrees (Kuz'mina and Lyapin 1980: 14, 20). 
Ceramics in central Kazakhstan were produced at 700 degrees (Kuznetsova and 
Teplovodskaya 1994: 127). As the temperature of a fire is 650/800 degrees, and 
in a kiln 900 degrees (Semenov and Korobkova 1983: 223, 224), it is possible to 
state that Andronovo vessels were fired in a deoxydizing atmosphere which 
conditioned its black-grey color (Avgustinik 1956: 152-153). It is quite possible 
that in order to obtain a beautiful black shade vessels were saturated with organic 
substances (e.g., a solution of flour, grass); pyrolusite and iron ore were added to 
the hearth (Ivanov 1979: 85). This explains iron ore finds in Andronovo settle- 
ments. From time to time Andronovans practiced oxidizing firing in an open kiln 
with sufficient access to oxygen so pottery would achieve a yellow-red color. 

A comparison of production technology and the ornaments on Petrovka and 
Alakul' ceramics indicates their distinct genetic relation. However, Alakul' pots 
differ in their standardization, thinner walls, richness of ornamental elements and 
their combinations, which points at considerable progress in the pottery of the 
Alakul' period. 

Fedorovo 

Fedorovo sites, dating to the 15th 13th centuries BC, are found in the forest- 
steppe, steppe, semi-desert and mountain zones: of the Urals, on the Tobol, in 
northern and central Kazakhstan and in Central Asia, where they coexist with 
Alakul'; they are also found in the eastern Andronovo range in Kirgizia, in 
valleys of the Tian-Shan, Pamirs and Tadhzikistan, in eastern Kazakhstan and 
along the Irtysh, Ob and Yenisey where Alakul' sites, other than the cemetery at 
Jermak, are unknown. Their southern border is Shortughai in northern 
Afghanistan (Francfort 1989, fig. 19-21, 41). The fabric of Fedorovo pots 
contains admixtures of sand, gravel, mica, rarely chamotte. The use of talc for 
temper, a characteristic of the Ural cultures from the Neolithic onwards, is absent 
from Fedorovo vessels of this region which indicates the alien nature of the 
Fedorovo complex in this area. Inside the vessels, mainly near the base and 
under the rim, there are traces of a small fingerprints. Forensic scientists have 
concluded that these were made by women. Vessels have thin walls, usually 6- 
8mm. Fedorovo ceramics are of a high quality: they are well formed, both the 
outer and inner surfaces are well-polished. The vessels were produced by spiral 



72 CHAPTER FIVE 

and ring building, the bands placed slanting to the outside; the shoulder of a 
Fedorovo vessel is rounded and sloping. Then the rim was attached. Fedorovo 
potters employed two methods for building a base. Where the base of the vessel 
was round, they were made in a pit beginning from the base (bottom start). A 
ring forming a ringed foot (la) or a small flat plate (lb) was attached and 
flattened out. This type of base-building with a round bottom is archaic and a 
Neolithic survival (Figs. 2; 11-13; 19-22). 

Another way of preparing a base in the Fedorovo tradition was to insert a 
separately modeled round bottom which was carefully attached to the walls from 
the inside. In this way the characteristic form of the Fedorovo vessel with its 
rounded shoulder and small-diameter base was a product of its ceramic-making 
technology. There are two types of Fedorovo pots. 

Type 1 vessels strictly adhere to a set of proportions. The rim diameter is 
equal to the height of the vessel, while the maximum width of the body is one 
third of its height. The rim diameter is usually equal to the maximum diameter of 
the body (variant la); rarely does the rim have a more cylindrical form where its 
diameter is less than that of the body (variant lb). The base diameter is less than 
half of the body diameter. The components of the vessel — body, neck and rim — 
are emphasized through zonal ornamentation, separated by division lines where 
the bands join; sometimes the foot of the base is emphasized by a flute from the 
outside. Pots of subtype I are characteristic of early Fedorovo sites; they date to 
the 15th 14th centuries BC on the basis of finds in closed contexts of the Alakul' 
and early Timber-grave (Pokrovskiy) complexes. 

Pots of subtype II show deviant proportions. They are more pot-bellied: the 
maximum width of the body is at the middle, the rim is sometimes poorly 
expressed, the base is wider and without a foot. The zonal character of the orna- 
ment is broken: in some cases the ornament is present over two zones, the bor- 
ders are moved below and the connection between the ornamental zones and the 
form of the vessel is not clear. In other cases the decoration is poorer; negative 
ornament is often used. Pots of subtype II are characteristic of Late Fedorovo 
complexes, dating to the 13th century BC according to finds from closed com- 
plexes alongside numerous metal objects of that period and Cherkaskul' vessels. 

Fedorovo potters employed a medium-sized and small-toothed stamp with 
rectangular, square, triangular, semi-oval or round teeth in section; comb rocker 
and caterpillar stamps appear in northern Kazakhstan from time to time. 

Bone, clay and bronze-toothed stamps are found on Fedorovo settlements 
(Sal'nikov 1951b: fig. 15,6; Chernikov 1960: fig. 13, 1,2, table 14, 2-7). Smooth 
stamps were rarely used. Flutes were used as independent elements of decora- 
tion; they covered the whole vessel or only its upper third; more often they divi- 
ded it into ornamental zones. Flutes were made with an astragalus with carved 
grooves. 

Fedorovo ornament was made over three zones. Initially, lines of division 
were imprinted, then the zones were ornamented without preliminary division 
inside the registers which seldom resulted in a break in rapport. Ornamental 
bands were drawn horizontally. 

M. P. Gryaznov suggested that the Andronovo ornamental system was trans- 
ferred onto the pots from multicolored textile patterns. Actually, many of the 
variants of Andronovo patterns can be made by moving and matching two 



CERAMICS 73 

opposite stripes with the same simple elements (Fig. 12). The original multicolor 
decoration was imitated on ceramics by either the background or the pattern of 
the lines. In modern mathematics we have several types of symmetrical compo- 
sitions: mirror, axis of several orders, etc. The Andronovans knew them all. 
Fedorovo carpet decoration was reproduced employing an oblique grid which 
conditioned the specific form of Fedorovo ornamental elements: oblique triangle, 
oblique swastika, oblique meander and its modifications, triangular scallops and 
hanging triangles (Chernetsov 1948: 151, 152; Zotova 1962: 177-180). Rows of 
oblique notches, horizontal and vertical herring-bone, sometimes reaching the 
base, were widely used reflecting Neolithic traditions on Fedorovo vessels. 

A comparison of Fedorovo and Alakul' pots reveals some important similar- 
ities in applying bands and similar proportions: division into three zones — rim, 
neck and shoulder with ornament over these three zones (Fig. 1 1). Almost all of 
the elements and most compositions of Fedorovo and Alakul' ceramics are uni- 
form and differ only with respect to their application onto a straight or oblique 
grid (Figs. 12; 13). The totality of these general principles of the Fedorovo and 
Alakul' types makes it possible to assign them to an Andronovo unity with a 
ceramic complex that is different from other Bronze Age cultures of the Eurasian 
steppes. The difference between Fedorovo and Alakul' ceramic complexes was 
conditioned by different technologies of forming and ornamenting a vessel, i.e., 
they reflect two various traditions of production. As domestic pottery skills are 
transferred within a kinship group and they are an important ethnic indicator, the 
existence of two technological traditions proves that two different, although 
related, ethnic components participated in the formation of the Andronovo unity. 
This eliminates a direct genetic connection between the Fedorovo and Alakul' 
people and presupposes their different origins. As mentioned above, Alakul' 
ceramics are connected with Petrovka and Sintashta and originate from the Early 
Bronze Age steppe cultures of Eastern Europe, including the Urals. The genesis 
of the Fedorovo type is not clear. In western Siberia Fedorovo sites replace 
Krotovo, Samus', Afanas'evo and Okunevo and they are not in any way 
genetically close; their interaction is reflected only in the appearance of specific 
ornamental motifs. The creation of the Fedorovo complex in the Urals is unlikely 
because the specifically Ural use of talc temper is absent from Fedorovo vessels. 

One working hypothesis is that the Fedorovo sites were established in central 
Kazakhstan on the basis of Neolithic sites such as those found by M. N. 
Klapchuk at Karaganda where there were stone tools, beads of ore, bones of 
(?domestic) animals, and flat -bottom ceramics comparable in some technological 
peculiarities to Fedorovo. A second possible center of formation for the Fedo- 
rovo type is eastern Kazakhstan where sites of Ust'-Bukon' and the genetically 
connected early Fedorovo Kanay type indicate a burial rite and ceramics with 
archaic features (Chernikov 1960: 16f, tables 6, 19, 3, 5; Kuz'mina 1994: 259f). 
N. A. Tkacheva (1997: 16) independently came to the same conclusions. 

In Kazakhstan one finds a specifically Fedorovo cremation rite. Finally, the 
territory of the Andronovo anthropological type, indicated in Fedorovo burials, is 
confined to northern, central and eastern Kazakhstan. Hence, the Fedorovo 
population probably migrated to the east, replacing in Siberia the genetically 
unconnected bearers of the native Okunevo, Krotovo, and Jelunino cultures. All 
of this presupposes a local Kazakhstan genesis for the Fedorovo complex. 



74 



CHAPTER FIVE 



The influence of the tradition of Andronovo Fedorovo ceramics is found in 
the forest cultures of the Final Bronze Age in the forest zone in the Urals and 
western Siberia on sites of the Cherkaskul'-Mezhovka, Pakhomovo, Suzgun, 
Elovka cultures, etc. (Sal'nikov 1964a; Obydennov 1981; 1986; 1997; Matveev 
1985; Potemkina et al. 1995; Korochkova 1987; Korochkova and Stefanov 1984; 
Shorin 1988, 1995; Moshinskaya 1957; Posrednikov 1973; Matyushchenko 
1974; Kosarev 1974; Kiryushin and Maloletko 1979). Vessels are of pot -like 
form with rounded shoulder, and the special band composition on an oblique grid 
reflects Fedorovo influence in the forest zone (see above). 

Mixed types 

In addition to pure Alakul' and Fedorovo sites, there are also mixed sites, dating 
to the 15th— 13th centuries BC, in Kazakhstan and Central Asia (see chapter 1). 
Ceramic assemblages that combine the elements of Fedorovo and Alakul' 
technology, form and mainly ornament, are characteristic of mixed complexes. 
Alakul' features prevail in all complexes except Tautary and Semirech'e. Thus it 
is possible to regard these mixed complexes as part of the Alakul' line of 
development. 

The syncretic types are characterized by: 1) a combination of vessels with 
rounded shoulder and vessels with a distinctly marked shoulder-join within a 
single complex; 2) different ways of fashioning the base; 3) a combination of 
ornaments over oblique or straight grids in a single complex or sometimes even 
on a single vessel; 4) a combination of typically Fedorovo and typically Alakul' 
decorative motifs (Figs. 11; 13; Map 11). 



Zone 




hi 
o 
et 
o 
< 


> 

I 


s 

o 


Q 

8" 

o 
< 

o 


19 rf 

< 1 

? 


H 

o 

E. 


F 

S 


Chelyabinsk 


22 


2 


12 





13 





49 


Uy-Uvel'ka 


7 


4 


6 





2 


3 


22 


Verkhny Tobol 


7 


1 


2 











10 


Magnitogorsk 


15 





3 








4 


22 


S 


W. Kazakhstan 
and Sol'-Iletsk 


20 


1 


6 


31 





2 


60 


N. Kazakhstan 


22 


4 


5 


4 


9 





44 


C. Kazakhstan 


18 





14 


22 


10 





64 


Kirgizia 


1 








8 


3 





12 


E. Kazakhstan 
and Irtysh 


9 











16 





25 


F 
S 


Ob 


11 











24 





35 


Yenisey 


6 








1 


21 





27 



Table 3: Correlation of site types 

Only sites with statistically analyzed ceramics have been counted. F = Forest, S = Steppe. 



CERAMICS 75 

Such mixed ceramics originate from sites which also display a mixing of the two 
burial traditions as well. In syncretic complexes of the Alakul' line of develop- 
ment there are local peculiarities. This permits us to identify sites of the Sol'- 
Iletsk type (Fig. 24), Kozhumberdy in western Kazakhstan (Figs. 23, 24) Aman- 
gel'dy in northern Kazakhstan, and Atasu in central Kazakhstan (Fig. 25). In 
southern Kazakhstan the Tautary cemetery must be of a special type, probably 
along the Fedorovo line of development (Fig. 27). There is also a syncretic 
Semirech'e type (Fig. 26: 1-14). The distribution of mixed type sites and a rough 
count of the number of different types (Table 3) indicates that they are spread 
over all of Kazakhstan and in Kirgizia and constitute more than half of all known 
Andronovo complexes. Many sites of Central Asia are also of mixed character. It 
speaks in favor of intensive assimilation and integration of the two initial 
components of Andronovo unity — Alakul' and Fedorovo; moreover, the strength 
of links within this unity is much higher than with other cultures. 

Alekseevka 

Sites of this type were identified by O. A. Krivtsova-Grakova (1948) and T. M. 
Potemkina (1975; 1985). Vessels of the Alekseevka type are present on 
settlements of the 13 th 10th centuries BC in the Urals, across all of Kazakhstan, 
the Altai and in Central Asia. Andronovo specialists tend to regard them as 
independent cultures: Zamaraevo, Alekseevka, Sargary, Begazy-Dandybay. K. 
V. Sal'nikov (1948b; 1951b) assigned all Ural ceramics of the late Bronze Age 
to the Zamaraevo stage of the Andronovo culture; A. M. Orazbaev (1959) and 
M. N. Komarova (1962) assigned them to a special Zamaraevo culture. V. S. 
Stokolos (1972) revealed the specific character of the Zamaraevo complex itself, 
and M. F. Obydennov (1986; 1997) stated that the Zamaraevo (Mezhovka) com- 
plexes originated on the basis of the Fedorovo-Cherkaskul' complexes. G. B. 
Zdanovich (1973a) separated the Zamaraevo and fl'inka ceramic types in nor- 
thern Kazakhstan. S. Ya. Zdanovich (1974b; 1979) made an attempt to identify a 
special Sargary culture. The Zdanovichs believed that the Sargary (Alekseevka) 
ceramics were either genetically connected with Fedorovo or appeared as result 
of a migration of late Timber-grave tribes. V. V. Evdokimov (1975a; 1984) 
undertook new excavations of Alekseevka settlements, isolated out a pure 
Alekseevka complex, and connected its genesis with Alakul'. 

T M. Potemkina (1975; 1979; 1985) divided the pottery of the Tobol into a 
Zamaraevo forest group and an Alekseevka steppe group, but there are many 
mistakes in her classification. 

The Alekseevka type of ceramics is found as far as the Altai steppe 
(Mogil'nikov 1976; Ivanov 1987; 1989; Udodov 1988; Kiryushin et al. 1990; 
Kiryushin and Luzin 1990) and archaeologists have suggested that the 
Alekseevka population came from the west. Settlements with a pure stratum of 
Alekseevka ceramics have also yielded imported ware of Namazga VI. 

What is the origin and chronology of the Alekseevka ceramics? 

The pottery contains the same tempering agents as Alakul': gravel, chamotte, 
large grains of sand, organic material and occasionally talc were added. The vari- 
ety of inclusions found in pots on the same site (Kuznetsova and Teplovodskaya 
1994: 159) reflects different technological traditions. Vessels are coil-built and 



76 CHAPTER FIVE 

the base is wide and the walls are applied from the outside as in Alakul' pots. 
Sometimes a textile mold has been employed (Alekseevka, Kayrak-Kum). There 
are vessels from central Kazakhstan that were produced by beating out the clay, 
which is typical for the Dandybay culture and reflect its influence (Kuznetsova 
and Teplovodskaya 1994: 129). The proportions are more elongated than those 
of the Alakul' vessels, the shoulder is smoothed, the rim is lower, sometimes 
yielding a complex profile. Firing was in the open, uneven, and often yellow or 
red in color. Zonal decoration is broken due to the change in the proportions of 
the vessel: it is found in two zones — on the rim and shoulder, sometimes only on 
the shoulder, where the semantically important decoration has been transferred. 
The ornament of Alekseevka ceramics is very poor: smooth stamp and incised 
ornament prevail as well as nail pinching and impressions; also large-toothed 
stamp is preserved. Only vertical and horizontal herring-bone and isosceles 
triangles, rarely rhombuses and simple meanders are preserved from the whole 
repertoire of Andronovo decoration; cross-forms also appear (Figs. 17; 18). 

The analysis of the technology and ornament of Alekseevka ceramics 
indicates that their origin is connected with Alakul' ceramics and mixed types on 
the Alakul' line of development. 

A characteristic feature of Alekseevka ceramics is the predominance of 
applied-roller decoration at the shoulder, sometimes with drooping tendrils. In 
the 13th 9th centuries BC this distinctive fashion spread over a vast territory 
from Asia Minor: Troy VIIc (Blegen 1958: 2 : 282, 284, 285), the Danubian 
Noua culture and late Timber-grave sites of the Ukraine, Don, Volga as far as the 
Urals and Kazakhstan, south Central Asia and Iran: Giyan 1 (Contenau et al. 
1935: 216, fig.; Kuz'mina 1967, Chernykh 1984: 246-258; Fig. 51). But this 
ornament was applied to vessels which universally preserve their local 
technology and form, thus one cannot distinguish a special 'roller ware culture'. 
A comparison of Alekseevka and late Timber-grave ceramics testifies to their 
great similarity which can be explained by the strengthening of contacts in the 
steppe following the transition to nomadic cattle-raising in the Final Bronze Age 
and the common trends in the development of two related ceramic traditions, 
Timber-grave and Alakul'. Their convergence was a result of the coarsening of 
the ceramics with the preservation of only the simplest decorative elements. 

Thus, the Alekseevka type of pottery should be seen as the final stage of the 
Alakul' line of development where there was a succession of production 
traditions and the presence of transitional complexes, e.g., the Alekseevka 
cemetery where Alakul' pottery and one modeled vessel with applied-roller on 
the shoulder co-exist (Krivtsova-Grakova 1948: fig. 54,1). 

The uniformity of the main forms and ornaments of such ceramics over the 
whole vast Andronovo region does not permit us to single out local Sargary, Tru- 
shnikovo and other cultures. All of these complexes are to be called Alekseevka 
according to the first site discovered. Moreover, it is not justified to combine 
such sites with Dandybay (Margulan 1979) or reconstruct a Zamaraevo-Begazy 
culture (Avanesova 1991), because Dandybay complexes are not genetically 
related to Andronovo complexes. The succession of Alekseevka and Alakul' and 
mixed Andronovo complexes is evident not only in pottery but also in burial rite, 
types of houses and settlements, clothes, ornaments, tools, and weapons, which 
prove their cultural unity. 



CERAMICS 77 

Specialists in the Volga Timber-grave culture have created a detailed division 
of Bronze Age sites into periods, defining a Period IV (Ivanovka) of the 12th- 
10th centuries BC and a Period V (Nur) of the 9th century BC (Agapov et al. 
1983; Kachalova 1989). 

Material from central Kazakhstan can also be divided into two stages: 
Alekseevka and later Dongal (Evdokimov and Loman 1982; Loman 1987; 
Varfolomeev 1987; 1988; 1991). The primary site is the settlement at Kent 
where a large area was excavated yielding an outstanding material complex that 
securely dates the Alekseevka type (Fig. 42). There are cheek-pieces of the 
Subbotovo and Boriyash types, perforated spearheads, bi-facial arrows, sickles 
of the Sosnova-Maza type, knives with ring tops, chisels, badge (Fig. 42: 3), a 
pair of compasses for making circles — types well-known in closed Eurasian 
burial complexes and hoards of the 12th (13th)-10th centuries BC. Wheel-made 
pottery of the later Namazga VI stage is also encountered. Metal objects and 
cheek-pieces confirm the proposed chronology. A razor from Sargary is especial- 
ly important as it sets the lower date of the complex to the 13th 12th centuries 
BC (Kozhomberdiev and Kuz'mina 1980). Thus, Alekseevka sites date to the 
13th 10th centuries BC. They are synchronous with Sabatinovka in the Ukraine 
and H'inka on the Volga both in respect to ceramics and other artifacts. 

Alekseevka-type pottery is also found at the settlements of Atasu, Myrzhik 
Buguly 2, Suukbulak, Ust'-Kenetay, Karkaralinsk, Tashik, Kopa 1, Akimbek, 
Upais, etc. In stratigraphy it overlays the Fedorovo and Alakul' and mainly the 
Atasu-type layer. Pure Alekseevka complexes are found in the cemeteries of 
Aydarly and Dermen; in some sites Alekseevka vessels are combined with 
Dandybay. 

Dongal 

The next stage is represented by Dongal sites (Loman 1987). These are the 
settlements of Dongal, Tagibay-Bulak, the upper layer of Kent, and Dzhazybay 
and Myrzhik. 

The ceramics are characterized by degeneration and the disappearance of 
stamped ornament, roller ornament becomes narrower and moves to the neck. 
Beads and appliques on the body of the pot are characteristic. Analogous pottery 
was justly classified as a separate type on the Trushnikovo settlement in eastern 
Kazakhstan (Chernikov 1960, Tkacheva 1997) The closest analogy is found 
among the Nur type ceramics on the Volga. It is known also in the sites of the 
Trans-Caspian and south Turkmenia (Kuz'mina 1988a). 

The Dongal type dates to the 9th century BC on the basis of the Trushnikovo 
and Kent complexes. This type fills the gap between the Andronovo and Saka- 
Sarmatian vessels and provides an origin for the latter. K. F. Smirnov (1964: 
112-127) concluded that "there is a direct dependence of a number of forms and 
ornaments of Sauromatian ceramics from the late Timber-grave/ Andronovan, at 
the same time production techniques preserve traditions of the local population 
of the Bronze Age." Sauromatian ceramics as well as Alakul' and Alekseevka 
were fashioned from clay with sand, gravel, chamotte temper (plus talc in the 
Urals) by the coil method of ring modeling on a wide protruding base; a join can 
be traced on the vessels; ornament was applied over two zones in flutes with 



78 CHAPTER FIVE 

tooth or smooth stamps as a zigzag, herring-bone, triangle, and rarely a rhombus. 
Sarmatian pottery develops the Sauromatian traditions; the vessel is made from a 
fabric-covered block mold, ornamentation was by applied-roller with a smooth 
or toothed stamp in zigzags, herring-bones, triangles (Voevodsky 1930: 65-66). 
Alakul' and Alekseevka pottery traditions are present within other groups of 
Iranian tribes of the Early Iron Age: the Wusun of Kazakhstan and the Saka of 
the Altai (Rudenko 1963: 90-91). 

Thus, the analysis of Andronovo ceramics leads us to the following 
conclusions: 

1 . With respect to ceramic technology, form of vessel, zonal and ornamental 
elements, and composition of fabric Petrovka ceramics are close to those of the 
Poltavka culture and the later Potapovka culture, i.e., the early stage of the 
Timber-grave type culture; the similarity of their ceramic traditions can be 
explained by their genetic relationship with Sintashta. The origins of this ceramic 
tradition are found in the Eneolithic of Eastern Europe. 

2. Alakul' ceramics developed directly on the basis of the Petrovka which 
reflects the ethnic unity of the Alakul' and Petrovka peoples. 

3. The ceramics of the Iranian Saka and Sauromatian cultures preserves the 
traditions of Andronovo pottery which leads us to regard the Iranian Sauro- 
matian-Sarmatians and Saka as descendents of the Andronovans. 



Dandybay 

Sites of the Dandybay culture date to the very end of the 1 1 th 9th centuries BC 
and they are distributed over central Kazakhstan. Individual Dandybay pots are 
found in the Kirgiz (Dzhail'ma, Kainda, a burial at Vorontsovo, in Semirech'e 
(Bien settlement), in Khorezm (Tagisken), in northern Kazakhstan (Sargary), in 
various sites of the northern Caspian in collections of V. D. Beletsky, A. N. 
Melent'ev (1972) and I, B. Vasil'ev (Vasil'ev et ah 1986) and from the Il'evskiy 
cemetery in the Volgograd region (Mamontov 1980: 158). 

Vessels have a globular body with a very small flattened base and cylindrical 
neck sometimes with an inverted rim. The sherds are thick, the paste contains a 
rather specific admixture of calcified bone and crushed granite (Kuznetsova and 
Teplovodskaya 1994: 128). The exterior surface is black, bright-yellow and 
dark-red in color and usually finely polished. Ornament is in thin, deeply 
imprinted tooth or smooth stamp; various figure stamps are widely used; nail 
impressions, relief, black and brown painting are also present. Decoration covers 
the whole body, the neck is without ornament or has a different decoration. 

Staggered oblique rhombuses, protruding opposed triangles, inscribed angles 
and broken shaded bands and also lattice shaded triangle and rhombus composi- 
tions situated over a grid are popular. Neither in form, decorative motifs nor 
manufacturing technique is the pottery similar to Alakul' or Fedorovo. Dandybay 
ceramics were produced by beating a whole piece of clay with a small round 
stone and a small spoon (Gryaznov 1952: 147). This technique was found in the 
Karasuk culture which also shares some similar ornamental motifs with Dandy- 
bay ceramics which prompted M. P. Gryaznov to assign sites of the Dandybay 
type to the Karasuk culture. A. H. Margulan opposed this (1979: 327-333) as he 
believed that the Begazy-Dandybay culture represented a direct continuation of 



CERAMICS 79 

Andronovo and was the prototype of the Saka. This conclusion is disputable. 
Taking into account the technological, formal and ornamental peculiarities of 
Dandybay ceramics and their relation to Karasuk, the Dandybay complex should 
be seen as a separate culture which is genetically unrelated to Andronovo. 

A ceramic technique similar to Dandybay is still employed by the Turkic 
peoples of Siberia — the Yakuts and Shorts (Podgorbunsky 1928). Craftswomen 
make a clay ball, press it into a hollow shape, and then beat out the body with a 
small spatula or wooden spoon (Voevodsky 1930: 60-61). This is a peculiar 
technique and is known in Siberia only among Turkic peoples which makes this 
technique of beating out the clay an important ethnic characteristic relating to the 
Turkish language family and suggests a probable migration from Siberia or 
Central Asia of the Dandybay peoples. 

There are two types of pottery coexisting in central Kazakhstan during the Fi- 
nal Bronze Age (Dandybay and Alekseevka). They are undoubtedly synchronous 
as they are recovered from the same closed complexes, e.g., in the cemeteries of 
Sangru 3 and Begazy, Aydarly, and Bola-Kulboldy (Margulan et al. 1966: 183-6, 
fig. 60,93, tables 1,19; 1979, fig. 29, 234, table 7). Aleeksevka pottery absolutely 
predominate on settlements of central Kazakhstan during the final Bronze Age, 
particularly at Kent (Varfolomeev 1991). Consequently, we see that the native 
Andronovo population dominated in central Kazakhstan during the 1 1 th— 9th cen- 
turies BC. It produced Alekseevka vessels, but a small Mongoloid group migra- 
ted from the east, penetrated into the area, and occupied a dominant position in 
the Andronovo environment. Slight traces of Mongoloid characteristics on the 
skulls of Europoid Saka (Ginzburg 1970: 199) are probably to be explained by 
the participation of a small intrusive group of Mongoloid Dandybay people who 
contributed to the ethnogenesis of the Saka of central Kazakhstan and Kirgizia. 
Huge tombs of the Begazy type were built for them, with ceremonial Dandybay 
pottery rarely encountered on settlements. The newcomers were soon absorbed 
by the Andronovans. The culture of the Saka was formed on a late Andronovan 
basis. 

Traditions of the Saka technique of hand-made pottery survive in remote 
districts of Tadzhikistan. The technique of ring modeling is encountered among 
the Karategin and Darvaz Tadzhiks, in Khuf, on the Zeravshan, Yagnob, Vakhsh 
and Kafirnigan (Semenov 1903: 39; Zelenin 1927: 99-100; Peshchereva 1929: 
27-28; 1959: 25-39). The author personally witnessed this method of ceramic 
production on the Vakhsh and Zeravshan. In Darvaz, at the settlement of 
Gumbulak in the Faizabad district and on the Vakhsh in Kangurt district and 
Baldjuan a vessel is made on an inverted pot or on a special mold of pressed 
dung with clay. A vessel or a block is covered with a wet cloth and the potter 
(female) fashions the lower part of a pot. After drying she puts the unfinished pot 
on a wooden dish and finishes modeling the rim (Ershov 1956: 7; Peshchereva 
1959: 28-29). In other cases, the walls of the vessel are modeled on a wide 
protruding base placed on a stone stand (Peshchereva 1959: 251). 

A vessel is made in three stages as was the case with the Andronovo 
ceramics: the lower part of the pot is called the bunuk, the body with its shoulder 
is the lona and the rim is the moruk. The pot is said to be made "with one wall", 
the carinated body of the vessel is made "with two walls", a tri-partite vessel 
with a rim is made "with three walls" (Peshchereva, 1959: 33, 35). The firing of 



80 CHAPTER FIVE 

the vessels among the Mountain Tadzhiks has been described by Grigor'ev 193 1: 
3 and Peshchereva 1959: 40-43, fig. 10. Women fence in a circular enclosure, the 
khumb, with stones, then put several layers of pressed cow dung (pur) over an 
area, 1.5-3m in diameter, on which the vessels are compactly placed, inverted or 
on their side, and they set twigs on fire; the pressed dung covers it closely from 
above, and this pile is fired for an entire night. In Tavil-Dara the firing is done in 
a special small pit; small vessels are sometimes fired in the domestic hearth. 

Similar pottery traditions are found in northwest Pakistan and Afghan Bada- 
khshan in the settlements of those small groups who speak a relict Indo-Iranian 
language. The population of northwest Pakistan is descended from the first Indo- 
Iranians who came into Hindustan from the homeland, partially before the 
division into separate Indian and Iranian branches, thereby preserving archaic 
features in their languages and mythology (Morgenstierne 1973; Jettmar 1975; 
Fussman 1977). The ceramic production of these peoples also reflects very 
ancient traditions. Ceramics were handmade. Clay, mixed with river quartz, sand 
and down, is crushed with a stone, mixed with water and kneaded. Then the base 
is formed on a flat wooden rotating saucer or dish (Rye and Evans 1976, tables 1, 
3, 34). Then a band is attached to it and after that the next band is inserted 
slanting inwards. The walls are made even with a rib or small spatula from the 
outside and with a small stone from the inside (Rye and Evans 1976, tables 20, 
21), then the surface is washed and polished with a small stone or egg-shaped 
piece of baked clay. The pots are fired in a pit, one meter in diameter (Rye and 
Evans 1976, tables 5, 10, 17, 21,63); its walls and base covered with stone tiles, 
sometimes coated with clay; in other cases a round hearth is floored with small 
irregular unbaked bricks and a dome is constructed on the top resembling an 
oven. A brick oven from the Andronovo settlement at Kipel' was probably of 
similar construction. All other types of hearths in Pakistan are analogous to those 
of the Andronovo culture. 

A similar ceramic technology was borne by the Aryas into India. E. A. 
Grantovsky (1981; 1998) was the first to study the ceramic characteristics as 
they related to the origins of the Indo-Iranians. He demonstrated that the Aryas 
did not know the pottery wheel in their homeland and thus disproved the 
hypothesis of some scholars who set their homeland in Iran. 

The evidence for ceramics in Vedic literature was collected by B. P. Sinha, in 
Brahmanic literature by M. Pandey (Sinha 1969: 155-160, 301-313). W. Rau 
(1972; 1974) analyzed all the data on pottery in Vedic India. The Upanishads 
and the Brahmanas distinguish between hand-made ceramics that were pleasing 
to the gods and made by those who performed Aryan sacrifices in the manner of 
the fathers, grandfathers (pitars) and those ceramics that were made with the 
help of a wheel by a potter (kulala-, a sudrd) who was not included in the Aryan 
community and did not participate in their sacrifices. Wheel-made pottery was 
not good for the sacrifice and belonged to the Asuras, the hostile aboriginal 
tribes and evil gods. "Asura vessel springs from the potters' hands, it is made by 
a potter on the wheel" (Maitrayani-Samhita 1.8.2-3; 2.9.5); "Some vessels are 
made on the wheel, some are made without. What is turned on the wheel belongs 
to the Asuras; what is made without a wheel belongs to the gods. Thus for the 
Agnihotra sacrifice a pot must be made without using a wheel" (Kathaka- 
Samhita 6.3; 17.13; Kapisthala-Katha-Samhita 4,2; 27,3). 



CERAMICS 8 1 

Similar descriptions of vessels fashioned for Aryan rituals are to be found in 
the Satapatha-Brdhmana (6.5.1.1-6.5.4.17; 14.1.2.9-25), in four Samhitas of the 
Black Yajurveda - the Maitrdyanl-Samhitd (3.1.6-8), Kathaka-Samhitd (19.5-7), 
Kapisthala-Katha-Samhitd (30.4,5), TaittirTya Samhitd (5.1.5,6) - as well as in 
Taittirlya-Aranyaka (5.2.8-5.3.9), Kdthaka-Brdhmana (93.13-96.13) and the 
White Yajurveda. They are the ukha-, mahdvTra-, kumbha-, sthdll-, etc. A vessel 
would be formed on the ground, or on a high round enclosed place covered with 
sand (pravargyd-). Water was added to the clay and mixed with five substances 
as "earth was created initially from five substances": ground sandstone, gravel, 
chamotte, goat hair, parts of some plants, and, following the Satapatha- 
Brdhmana, goat milk and plant resin. Then the Aryan wife made a clay tile with 
her hands — a foot in size — to serve as the vessel's base. Then the Aryan himself 
made clay tiles decorated with three signs and formed a vessel. According to the 
Satapatha-Brdhmana he forms the flat bottom of a vessel (ukha-), "then turns up 
an edge. Then he puts in place the first clay band... After he presses this band 
and moistens it well he puts on another band", he forms a three band ukha- using 
the divine measures of a span in height, a span in width", "he forms the ukha- 
from inside and from the outside" and "polishes it with a bunch of grass". We 
have a detailed description of coil modeling technique from three bands begun at 
the bottom. The Maitrdyanl-Samhitd (3.1.6-8) and Kathaka-Samhitd (19.5-7) 
also prescribe: "From three bands the ukha- is to be made". The Aryans formed 
the kumbhd-, mahdvTra- vessels, and tryuddhi- cauldron (meaning "made of 
three bands") from three coiled bands. 

Neither color-slipping nor painting the vessel is mentioned in Vedic texts. A 
bamboo stick (stamp) appears in the Taittirlya-Aranyaka for imprinting stamped 
decoration (Rau 1972: 1, 14). In the Satapatha-Brdhmana and other texts the 
ornamenting of the ukha- with modeled cones (nipples) and rolled relief is 
mentioned (Rau 1972: 1, 46). 

The modeled vessel was sun-dried and smoked with horse dung. Then the 
Aryan potter dug a pit oriented to the four cardinal directions. The potter put clay 
tiles and fuel into it adding grass. The fire was started during the day and the 
vessel was removed only on the next day. Then the pot was cleaned from the 
ashes and filled with goat's milk for cooling. This is a description of low temper- 
ature reduction firing. 

There is no doubt that the production of hand-made cult vessels and firing 
without a hearth could appear only among a people who produced all their wares 
without a wheel and whose ritual preserved memories of an ancient tradition. 

However, among the Andronovans and Tadzhiks pottery was made by 
women and in Vedic India it was made by men. Pottery was probably also a fe- 
male occupation among the ancestors of the Aryans. This is indicated by the fact 
that the priest's wife participated in the production of cult vessels. The Mother- 
Earth primogenitrix of all beings, the goddess Aditi and other goddesses, helped 
the Arya. "Aditi is Earth. With Aditi 's help he digs the earth, not to harm it; with 
Aditi's help he forms the ukha-"; "Dhisana is knowledge, the goddess... must 
light you in the house of Earth;" "goddesses must burn you;" "VarutrT is night 
and day... VarutrT — two goddesses must burn you, ukha-. They burn throughout 
the whole day and night;" "the wives of the gods' wives first made the ukha-" 
(Maitrdyanl-Samhitd 3.1.6-8; texts of the Kathaka-Samhitd 19.5-7; TaittirTya 



82 CHAPTER FIVE 

Samhita 5.1.6-7 are similar). In the Satapatha-Brahmana (6.5.1-4) it is said that 
"Great Aditi forms the ukha- skillfully, with force, using two hands". 

Vedic data are important for reconstructing complex ideological statements 
accompanying the industrial process. In the Atharvaveda (18.4.30) a vessel is 
identified with Aditi. In the Satapatha-Brahmana (14.1.2.9) it is said that "clay 
is Earth, water is sky. Mahavird- is made from clay and water". Ukha- is com- 
pared with Earth {Satapatha-Brahmana 6.5.1-4): "He makes ukha- as large as 
this Earth was first made...", "You are Earth, the base of ukha- (and the near- 
bottom part of the vessel - E.K.) is an earth living space" (it is connected with 
the Vasu gods), the second band of the vessel "is an air space" (it is connected 
with the Rudras), the upper band "is sky" (it is connected with the supreme gods 
Adityas), the walls of a vessel are parts of the world correlated to the gods, fa- 
vorable to the people". Identification of the three zones of a vessel with the three 
spheres of the universe is stressed in the Yojurveda. This idea is important for 
interpreting the semantics of Scythian artifacts including the Chertomlyk vase 
with its three spheres of the universe (Kuz'mina 1976c). The fashioning of a pot 
is compared to the act of creation, the production of each section is followed by a 
spell: "Rise! Become firm! Be big! Stand straight! You are stable, you are on a 
solid base". According to the Satapatha-Brahmana (6.5.4.17) and the Sukla- 
Yajurveda-Samhita (11,59) the applied-roller is the belt of Aditi, a cord handed 
by Varuna for sacrificing; this cord is modeled on the upper third of the body of 
the ukha-. These are "parts of the world", from it four clay borders come down 
vertically and terminate in modeled bosses: "gods forming the ukha-, these living 
spaces, milked all their wishes by these nipples, ukha- with four teats — it is a 
cow with four teats". The vessel is Makha — sacrifice, this is a sacrificed cow, 
this is a victim's head. The sacrificial vessel is dedicated to the god Mitra, 
"Owner of peoples," "Guardian of living spaces". The priest "obtains a wealth of 
descendants, possession of cows, good man's power, kinsmen" {Satapatha-Brah- 
mana 6.5.1-4). Prayers are sung while mixing the clay, burning a fire, placing a 
vessel. Varuna, Agni, Vayu, and Savitar are mentioned alongside with Mitra — 
ancient, partially Indo-Iranian gods. This reflects the deep antiquity of the very 
formation of the tradition. In highland Tadzhikistan the production of pots was 
also followed by complex rituals and spells; the protectress of craftswoman was 
called "mother" {momo), cf Aditi (Grigor'ev 1931: 2-4; Peshchereva 1959: 116- 
129). 

The Indo-Iranian origins of Vedic pottery are also supported by linguistic 
data: The Sanskrit name of the vessel kumbhd- exactly corresponds to Avesta 
xumba-, Tadzhik hum, Yagnob humb which means "vessel" and "enclosed 
ground for firing ceramics", Sanskrit kulala- "potter" corresponds to Tadzhik 
kalal "potter"; Tadzhik bunuk 'the lower part of a vessel with its base" goes with 
Avestan buna- "bottom, base" and corresponds to Sanskrit bundhya- "lower 
world". Thus comparison of linguistic and ethnographic data with evidence from 
Vedic literature and archaelogical materials permits us to relate Vedic ceramic 
production technology with that of modern northern Pakistan and the Mountain 
Tadjiks and it reveals the genetic proximity with the more ancient pottery of the 
Andronovo tribes. 

The technological process described in Vedic literature corresponds with the 
archaelogically reconstructed technology of Andronovo pottery. The most 



CERAMICS 83 

important details include the composition of the clay paste and the use of 
tempering (quartz, chamotte, plants, down), forming the vessel on a tile stand, 
building from the base, the ring modeling technique, polishing with grass, 
applied-roller ornament, firing in tiled pit, adding grass to the fuel. The most 
important features are the specific ring modeling and the three -part vertical 
structure of the vessel that permit us to compare Vedic pottery not just with 
Eurasian steppe pottery in general, but specifically with Andronovo pottery. 

Modeled vertical rollers and small bosses are found on Andronovo vessels; 
its different variants are present in the Sintashta and other Novokumak wares 
that are related to the Catacomb and Multi-roller ceramics (Smirnov and 
Kuz'mina 1977, fig. 9) and is met on 12th-9th century BC ceramics decorated 
by applied-roller with drooping tendrils (Figs. 17, 18). The many-sided vessels 
mentioned in Vedic literature are of great importance. In the Kathaka-Samhita 
(19.5-7), Kapisthala-Katha-Samhita (30.3-5), Taittiriya Samhita (5.1.6,7) the 
ukha- is said to be made with four, six, eight or nine-edges for magic reasons 
"against sorcery". Such square vessels are known only in Andronovo sites of the 
Fedorovo type (Maksimenkov 1978: tables 9, 5, 14-16, 22, 8, 9, 2, 3; Figs. 21:2, 
13, 16). These analogies are so specific that could not originate convergently. 

Ethnographers have observed that traditions of ceramic production are trans- 
ferred from mother to daughter within a community (Peshchereva 1959: 20; 
Sayko 1982: 15). Thus pottery technology is a solid ethnic indicator. Hence the 
evident closeness of Vedic and Andronovo pottery, the successive character of 
production from modern Indo-Iranians of Pakistan and Iranian Tadzhiks back to 
the Iranian Wusun, Saka and Sarmatians and finally to the Andronovans are 
extremely important arguments in favor of their Indo-Iranian identity. When the 
pottery tradition reached the Eneolithic of the Eastern European steppes, this 
specific archaeological category (ceramics) is rather important for viewing the 
Indo-European problem. T. V. Gamkrelidze and V. V. Ivanov (1984) propose a 
hypothesis that different Indo -Europeans groups arrived from the Near East 
through Central Asia in the second half of the 2nd millennium BC and Iranian 
tribes came only in the 1st century BC. Ceramic analysis does not support this 
hypothesis. 

There were two large regions within Eurasia during the Neolithic and Bronze 
Age: the zone of farming cultures of the Near East and Central Asia and the 
stockbreeding cultures of the Eurasian steppes and Central Europe. The culture 
of the ancient farmers of the Near East, Iran, India and south Central Asia 
developed in different ways from that of the Central European zone (Table 9). In 
Mesopotamia, Iran, and Baluchistan ceramic production had already progressed 
greatly by the middle of the 4th millennium BC: the potter's wheel was invented, 
and specialized craftsmen engaged in making pottery for market purposes had 
appeared (Sayko 1982: 92, 166). In southern Central Asia in the Anau culture 
this innovation belongs to the end of the 3rd millennium BC and beginning of 
the 2nd millennium BC, to the time of the Namazga IV/V transition (Masson 
1956: 295-309; Sayko 1971: 85-90; 1982: 91), in India to the time of the 
formation of the Harappan civilization. In contrast, in the stockbreeding zone of 
the Eurasian steppes hand-made pottery without a wheel and manufactured by 
women survived through the entire Neolithic and Bronze Age. 



84 CHAPTER FIVE 

Recently great attention has been paid to the pyrotechnological potential of 
society, i.e., ascertaining the maximum temperatures used by a given culture for 
their industrial processes (Ivanov 1979). The pyrotechnology of Eurasian steppe 
tribes who fired their ceramics and cast bronze in fires and hearth pits did not 
exceed 700-950° C. On the other hand, representatives of the ancient farming 
cultures of the Near East, Central Asia and India learned to make special two- 
tiered furnaces and kilns with high temperature for firing pottery in the 
Eneolithic. Their construction in the Bronze Age was perfect (Sarianidi 1958; 
1963; Masimov 1972; 1976). It is interesting to note that in the Avesta there is a 
word tanura- to denote an oven for firing ceramics. According to E. Herzfeld 
(1941) this word is not originally Iranian but was borrowed from Sumerian 
through Akkadian or another Semitic language. 

An important pattern in the distribution of different principles of 
ornamentation was already observed by M. V. Voevodsky (1936: 68-73): in the 
zone of ancient southern farming cultures painted ornament was universally 
predominant. The earliest pottery of the Near East and south Central Asia, Iran 
and India were as a rule color-slipped and painted (the first samples of painted 
wares appeared in Jarmo, Hassuna and Tepe-Sarab). The transition to mass 
specialized production marks the disappearance of painted wares, although local- 
ly painted or slipped wares can be found. In contrast, in the stockraising steppe 
zone there is no painting; here stamped decoration is to be found and in the Final 
Bronze Age applied-roller ornament appears as well. The percentage of ornam- 
ented pots decreases. It is necessary to note in this connection that in Vedic texts 
ceramics with stamped ornamentation are rarely mentioned. Andronovan ornam- 
ent is preserved in clothes, in the textiles of some ethnic groups of North-West 
Hindustan and among the Mountain Tadzhiks in particular (Bobrinsky 1900). 

Analysis has shown that in the sphere of ceramic production in the Eurasian 
steppes during the whole of the Bronze Age (17th [or 21st according to 14 C]- 
14th centuries BC) there is no influence of Near Eastern pottery and no break in 
the development of the Eneolithic tradition. On the contrary, in southern Central 
Asia we can trace the penetration of several groups of steppe herdsmen who 
brought hand-made pottery into the world of the ancient farmers. All this makes 
it possible to reject the hypothesis that a Near Eastern population migrated to the 
Eurasian steppes and claim rather that the steppe tribes moved in the opposite 
direction. 



CHAPTER SIX 

MINING, METALLURGY AND THE METAL INDUSTRY 

Mining comprises the extraction and refining of ore; metallurgy involves the 
smelting of metal from ores; the processing of metal includes the casting and 
forging of metal objects. 

Numerous deposits of ores of copper, gold, and cassiterite, from which tin is 
acquired, are known across the territory of the Andronovo culture. At some 
mines there is evidence for ancient workings but only a few of them can be 
confidently assigned to the Andronovo culture. Deposits suitable for exploitation 
in the Bronze Age must display two features: they must involve mineral sources 
that outcrop onto the surface to allow easy prospection and discovery; they must 
also consist of oxidized ores that were not too difficult to smelt. The mining of 
sulphurous ores was originally impossible as it presupposes a smelting 
temperature of 1300 degrees which was above the thermal potential achieved by 
Andronovans and became possible only in the Final Bronze Age. We attribute a 
mine to the Andronovo culture on the basis of Andronovo ceramics or tools in an 
ancient quarry or dump. There are also indirect but reasonable arguments for 
establishing the Andronovo age of a mine on the basis of its ore composition and 
the presence of ore and smelted metal on an Andronovo settlement. 

Descriptions of ancient mines have been made by travellers and geologists 
from the 18th century. Reports of ancient mines were compiled by M. P. 
Gryaznov (1935) for the Altai, M. E. Masson (1930; 1930b; 1934; 1936; 1953), 
B. A. Litvinsky (1950) and O. I. Islamov (1955) for Central Asia, V. T Surgay 
(1951) for Kirgizia. A new stage in the study of ancient mining was the work of 
the geologists V. A. Pazukhin (1926), B. M. Chudinov (1936), K. I. Satpaev 
(1941), L. P. Levitsky (1941), F. V. Chukhrov (1950), D. Khayrutdinov (1955) 
and especially N. V. Valukinsky (1948; 1950) in Kazakhstan. They not only 
described mines and collected chance finds but they also examined the settle- 
ments of neighboring miners. Their data are summarized by A. Kh. Margulan 
(1970: 3-30; 1973: 3-24). 

S. S. Chernikov conducted important research into Andronovo mining in 
1935 and 1937 in eastern Kazakhstan, and then again in 1938 in northern Kaz- 
akhstan (Chernikov 1948: 13-33; 1949; 1960: 118-136). He studied prehistoric 
workings and discovered the settlements of neighboring metallurgists. The large 
material recovered from the sites permitted him to date some of the deposits and 
reconstruct the process of ore extraction and processing; for the first time there 
were comparisons between the chemical analysis of the ores and metal artifacts. 

N. P. Kiparisova (Sal'nikov 1967: 275-277, fig. 41) examined several ancient 
mines in the Urals in 1950-1954. In 1949 A. A. Formozov (1951a: 118-119) 
examined the mines at Elenovka and Ushkatta which had been discovered by 
geologists and the nearby miners' settlements of the Elenovka microdistrict. 



86 CHAPTER SIX 

From 1959 to 1967 this metallurgical center was under systematic study 
(Kuz'mina 1962a; 1962b; 1963a; 1963b; 1964a; 1964b; 1965a; AO 1966: 113- 
114) and settlements and related burials were excavated; numerous traces of 
metallurgical and metal processing were revealed. 

The work of E. N. Chernykh (1970) marks a new stage in the study of 
Andronovo metallurgy. In 1967-1968 he examined the ancient ore sources of the 
Urals and studied the composition of ancient metal artifacts in comparison with 
ores from known deposits, defining metal groups according to their chemistry, 
and connecting some of them with ancient mines. I. V. Bogdanova-Berezovskaya 
began the study of the chemical composition of metals from Central Asia. 
Metallurgical and metal processing centers in Central Asia were established on 
the basis of typological classification of the metal objects and the composition of 
their metal (Kuz'mina 1966; 1967; Kuz'mina 1991). 

The discovery of the Sintashta settlements of metallurgists in the Urals was a 
major event in Andronovo research. Two ancient mines have been found so far: 
Vorovskaya Yama, which provided the ore for the settlement at Kuysak (Zaykov 
1993: 151; Zaykov et al. 1995) and Kisenet which is connected with the settle- 
ment at Ust'e (Vinogradov 1995: 18). It is likely that ore from sandstones of 
Tash-Kazgan was used; this contained an arsenic admixture characteristic of 
some of the metal objects from Sintashta. It is not clear how the Sintashta 
population exploited the rich deposit at Kargaly (Chernykh 1996; 1997). 

The unprecedented number of copper objects in burials and evidence of metal 
processing in every house in the settlement reveals the development of metallur- 
gy and metal processing. Several types of hearths were used for smelting ore: a 
grooved hearth, paved with stone; a two-chambered hearth and domed oven with 
flue connected with a pit for the disposal of ash. These constructions reveal an 
important advance in metallurgical development in comparison with the earlier 
Pit-grave period, because they raised the thermal potential of society. L. White 
regarded the development of thermal potential as a decisive factor in cultural 
history. 

The miners exploited the rich oxidized ore from the upper parts of serpentine 
deposits (Grigor'ev S. A. 1988, 1994; Zaykov et al. 1999). Slag containing 
increased amounts of arsenic suggests not only the exploitation of the Tash- 
Kazgan ore deposit but also artificial copper alloying, i.e., the creation of 
arsenical bronze, an important innovation in comparison with the Pit-grave 
period. Metal -working at Sintashta preserved its community character; there was 
no evidence for specialized skilled craft production. A larger portion of the metal 
objects from Arkaim were made of pure copper, some were made of copper with 
arsenic. In Sintashta 48% of the artifacts were made from copper with an 
artificial arsenic admixture, 34% was of copper with a natural arsenic admixture. 
In Kuysak two objects were made of pure copper, and one knife of tin bronze 
(Zaykov et al. 1999: 194-195). East European Abashevo metallurgy stimulated 
the formation of Sintashta metallurgy. 

Sintashta metallurgy is of a domestic nature, which is indicated by its traces 
in every settlement and house and the lack of any features of craft specialization. 
However, the volume of production was such that the Sintashta community as a 
whole (and not through specialized craft production), exported metal to the 
regions of the Volga and Don, areas poor in natural resources. Petrovka and 



MINING, METALLURGY AND THE METAL INDUSTRY 87 

Alakul' metallurgy inherited and developed the pre-Sintashta traditions 
(Grigor'ev 1995a: 122-126). 

V. D. Ruzanov (1982; 1987) studied the bronzes of the Chust culture. A. D. 
Degtyareva (1985) examined the bronzes of Kazakhstan and Semirech'e. The 
advantage of her work is in the analysis of technological processes. Detailed 
studies on ancient mines and settlements of metallurgists at Atasu and Myrzhik 
in central Kazakhstan (Alekseev and Kuznetsova 1980; 1983; Kuznetsova 1987; 
1989a; 1989b; Kuznetsova and Belovodskaya 1994; Kadyrbaev 1983; Kadyr- 
baev and Kurmankulov 1992; Zhauymbaev 1984a; 1984b; 1987; Kurmankulov 
1988) are of special importance for the study of the developed Andronovo 
metallurgy. 

A number of Andronovo copper mines were found by geologists and archae- 
ologists. These include Elenovka and Ushkatta (chemical group EU) in the west; 
their production reached 50.43% of west Andronovo metallurgy and was expor- 
ted westward where it comprised 24.6% of the metal of the Timber-grave culture 
(Chernykh 1970: 38, 40, table 5, fig. 40). The antiquity of these ancient mines is 
proved, in the first place, by the discovery of Andronovo ceramics and stone 
tools near the quarries and working areas; in the second place, by the discovery 
of ore fragments, slag, beads of copper and bronze objects made from Elenovka 
ore according to E. N. Chernykh (Kuz'mina 1962a; 1962b; 1963a; 1963b; 1964); 
in the third place, by the fact that the enclosures of the Andronovo cemetery at 
Elenovka are made of stone excavated from the nearest quarry (Formozov 
1951a: 118-119; Kuz'mina 1962a; 1962b; 1963d: 129; 1964b; Sal'nikov 1967: 
275). The Andronovans also exploited the southern Urals deposits of Bakr- 
Uzyak at Magnitogorsk, the group of Uchaly deposits by Uy-Tash-Kazgan 
(chemical group TK). The deposits of the Zau group were also exploited: Nikol'- 
skoe, Polyakovka, Narali, Voznesenskoe, Mednaya Gora, Kichiginskoe (Usf- 
Kaban). Ore from the Uvel'ka river deposit was found on the Andronovo settle- 
ment of Chernyaki 3 (Chernykh 1970: 40-45, 3, fig. 32). Andronovans also 
developed the Kargalinsk (Chernykh 1970: 48, 199) and Sol'-Iletsk deposits, 
where copper bars were found alongside Andronovo pots (Popov 1964: 262, fig. 
46). Ancient ore workings at Uro-Tyube and Yashilly were discovered, along 
with a stone hammer and miner's pick analogous to tools from neighboring 
Andronovo settlements (Chernikov 1948: 20,21,28). Central Kazakhstan was the 
largest center of copper mining with the richest oxidized polymetal ore deposits 
where one can find pieces of native ore weighing several tons (Pazukhin 1926; 
Satpaev 1929; Valukinsky 1948; Chukhrov 1950). The exploitation of mines in 
Dzhazkazgan — Kresto, Petro, Zlatoust, on the Dzhezdy river, etc. — extend back 
to the Andronovo period where numerous stone tools and ceramics fragments are 
found and near which the settlements of Andronovo metallurgists such as Mily- 
kuduk, Aynakol', Sorkuduk, Kulman (with numerous mining tools and evidence 
of metal processing) are situated (Satpaev 1929a; Valukinsky 1950; Margulan 
1973: 3-24; 1979: 233-254; Zhauymbaev 1984a; Kuznetsova 1989a; 1989b). The 
majority of ceramics in the Dzhazkazgan settlements belongs to the Alekseevka 
type, which dates the floruit of this center to the last quarter of the 2nd millen- 
nium BC. The exploitation of the rich Kenkazgan polymetal deposit began in the 
Andronovo period, which is demon-strated by the discovery of stone mining 
tools and Andronovo ceramics in quarry dumps and the similarity of the 



88 CHAPTER SIX 

chemical composition of copper from Kenkazgan ore with that of slag and metal 
objects from the large settlement of Atasu which lies 80km from the deposit and 
probably served as its base (Alekseev and Kuznetsova 1980; 1983; Kuznetsova 
1987; Kuznetsova and Belovodskaya 1994; Kadyrbaev 1983; Kadyrbaev and 
Kurmankulov 1994). Another ore source for Atasu was the mine at Sarybulak, 
20km west of the settlement. Slag, stone tools and ceramics were found near the 
quarry (Zhauymbaev 1984a: 114-117, fig. 1; 1984b; 1987: 109). The Andronovo 
age of the Altyn-Tyube deposit is unquestionable: its ore is close in composition 
to some imported Timber-grave objects (Chernykh 1970: 17). A settlement and 
Bronze Age stone enclosure lie near the Altyn-Tyube mine, where ceramics were 
found (Zhauymbaev 1984a: 117-119, fig. 2, 3; 1987: 109, 110). A large group of 
ancient mines is concentrated in the Karkaralinsk district: Meizek, Zheradur, 
Syrymbet, Kalmaktas, the latter of which is rich in native copper; copper blocks 
of 600-700kg are found there (Chukhrov 1950: 50; Margulan 1972: 5). Highly 
probable is the exploitation of the mines on the Tokraun river in the north near 
Lake Balkhash: Kenely, Sorkuduk and Kayraktas; this is demonstrated by a 
nearby concentration of settlements of Andronovo miners (Khayrutdinov 1955; 
Margulan, 1972: 18). Near Balkhash at the Tesik-Tas mine stone hammers and 
wedges, as well as Alekseevka type ceramics, were found (Zhauymbaev 1987: 
110). Several dozens of ancient copper mines are known from eastern Kazakh- 
stan. Andronovans extracted ore from the Karchiga deposit, where stone and 
bronze tools and ceramics were found (Chernikov 1949: 38-39; 1960: 118). In 
western Siberia, apart from the native copper deposits of the Altai, we also have 
the Kuznetsk Alatau deposit with large nuggets; native ore is also found in the 
Tom basin (Kosarev 1974: 24) but there is no proof that it was exploited by the 
Andronovans. Copper deposits are also found along the Yenisey where ancient 
mines have been discovered (Lev 1934), but definite Andronovo sites there are 
not known to the author. Andronovans also utilized sources in Central Asia. In 
the central Kyzylkum, in the Bukan-tau and Tamdy-tau mountains ancient mines 
have been discovered, and close to them in Beshbulak and Munbulak there are 
metallurgical sites with traces of copper smelting and slag, as well as Tazabag- 
yab, Alakul' and Fedorovo ceramics, which prove that Andronovans exploited 
these deposits (Itina 1961: 84; 1977: 136). An important copper extraction area 
was in the Nurata mountains north of the the Zeravshan: Nurata, Lyangar. Poly- 
metal and oxidized ore deposits and numerous ancient mines are known from the 
Karamazar mountains (Litvinsky 1963: 170): Uchkatly-Miskan, Kansay, Adras- 
man (a flint spear that dates from the beginning of its exploitation during the 
Bronze Age was found on the latter site). From the Almalyk mines (Masson 
1936) native copper and polymetal ores were extracted; that this occurred during 
the Bronze Age is shown by the discovery of a copper knife and a flexed burial. 
Copper -bearing sandstones and native copper deposits in Fergana are found in 
Naukat. It is likely that metallurgists from sites on the Kayrak-Kum worked on 
these, which is demonstrated by the similarity between Kayrak-Kum bronze 
objects and the ore from this deposit (Litvinsky 1963: 170; Kuz'mina 1966: 
analysis #79). Mines with ancient workings are known in Semirech'e: on the 
Chu river, in the Talas valley (Aktash), near Ketmen' -Tyube, on the Issyk-Kul', 
in the districts of Alma-Ata and Arpa (Masson 1930a: 44; 1980b: 35; 1936: 12). 



MINING, METALLURGY AND THE METAL INDUSTRY 89 

Of principal importance for the development of Andronovo metallurgy was 
the fact that the Andronovo culture was the main, if not the only, tin supplier on 
the Eurasian steppes. Cassiterite deposits for tin extraction are known in central 
Kazakhstan. The Atasu deposit was a tin source for the Atasu settlement, which 
is indicated by diagnostic admixtures in the ore (Kuznetsova 1987: 44). Ancient 
metal-working sites are also known at the Kalay-Kazgan mine in north Betpak- 
Dala (Margulan et al. 1966: 269; 1973: 5), on the Ishim, and in the Kokchetav 
mountains (Margulan 1972: 25). The richest tin mines are in eastern Kazakhstan 
in the Kalba and Narym mountains. The earliest site known is the mine at Myn- 
chunkur (shown by Andronovo ceramics, a stone mold for a knife and chisel), 
Cherdoyak (Andronovo ceramics and stone mining tools), Karagoyn (ceramics 
and two knives of the late Bronze Age) and others (Chernikov 1949: 10-36, 
tables vii, xii, 3, 4, xiii). There are subterranian placer cassiterite deposits on the 
Irtysh (Chernikov 1960: 135). Several deposits of placer cassiterite were recor- 
ded for the basin of the Tom (Kosarev 1974: 24). Rich cassiterite beds are in 
Central Asia: in the central Tian-Shan and in the Issyk-Kul' area, in the Zerav- 
shan ridge — Takfan; in the Zerabulak mountains near Samarkand — Changali and 
Kochkarly, and an especially rich deposit at Karnab west of Samarkand with 
many stone mining tools from the Bronze Age (Litvinsky 1950: 1954). Tin 
deposits permitted bronze casting to flourish among the Andronovans, helped to 
establish their active contacts with other tribes, and provided them with an 
outstanding role in the steppes. 

The Andronovans also exploited gold deposits, both ore and placers. In the 
Urals, in Bashkiria, it was extracted from the Kuseevskiy mine where polished 
stone tools have been found (Sal'nikov 1967: 278-279). North Kazakhstan was 
immensely rich in gold (Chukhrov 1950: 17-25). Placer gold was extracted at 
Borovoe and there were mines in the Stepnyak area (Stalinskiy, Bes-Tyube, 
Aul'naya Ploshchad'). Their undoubtedly Andronovan age is shown by the fact 
that settlements, where we recover Andronovo and mainly Alekseevka ceramics, 
metal objects and stone tools, are close to the metal working sites (Chudinov 
1936: 37-40; Chernikov 1948: 14-19). There are also gold veins and nuggets in 
central Kazakhstan. A large number of ancient mines are known here: in the 
Karkaralinsk district (Altynsu, Alabuga, Kyzyl -Espe, Akchagyl, Akzhal, Murza- 
Shoku) in the Boyan-Aul district (Altyn-Kazgan, Altyntas); in the Karaganda 
district (Zhosaly, Kushoku, Kenshoku); in the Ulutau mountains (Akshoku, 
Sorkuduk, Koskol, Obaly); in the vicinity of Lake Balkhash (Sayak) and many 
others (Chukhrov 1950: 4, 54; Margulan 1972: 17-18). Their exploitation by 
Andronovans is documented by stone tools in the mines and the location of 
settlements near the mines. Gold ore was extracted in eastern Kazakhstan at 
Kazanchunkur, placers were exploited on the Irtysh, shown by findings of nume- 
rous Andronovo bronze objects (Gryaznov 1935: 192-193; Chernikov 1960: 118- 
1 19). Central Asia is rich in gold. Placer gold was washed in the Fergana on the 
Sokh, Kassansay at Uzun-Akhmat (where a bronze knife was found), on the 
Naryn (a bronze axe was discovered in Uch-Kurgan). In the mountains near 
Tashkent placer gold was extracted on the Angren and Chirchik rivers. The 
antiquity of the gold-panning is proved by the Chimbaylyk hoard in a gold- 
bearing layer and the Andronovo burial at Iskander on the Chirchik near the 
deposit. Much gold is contained in the Zeravshan, Vakhsh and Kafirnigan rivers 



90 CHAPTER SIX 

that flow from the Pamirs. Written sources refer to its extraction in antiquity, but 
gold was probably extracted as early as the Bronze Age; from this period derives 
a hoard of an axe, chisel and spear found in a mine at Darvaz and the etymology 
of the ancient Iranian name of the Zeravshan river is "scattering gold". 

The Andronovans also extracted silver and lead. They were used, for 
example, in Semirech'e: at Tash-Tyube silver temple rings and lead beads were 
found. Sometimes lead was artificially added to copper. With regard to silver- 
lead mines there are ancient workings in central Kazakhstan at Berkkara and 
Kyzyl -Espe (Margulan 1972: 29-30), but it is not known if they were used by 
Andronovans or not. Rich polymetal deposits with much silver and lead are in 
Central Asia: Lashkerek (Islamov 1960) and Karamazar, where stone mining 
tools including "hoes" of Andronovo type were found in the Kansay mines. A 
copper knife was discovered in the Dzhol-say silver-lead mine in Kirgizia 
(Islamov 1960: 188). There are gelenite deposits (from which lead is extracted) 
in Talas Alatau (Litvinsky 1954: 24-25) where a copper knife and antler pick-ax 
were recovered. 

The Andronovans also used antimony, especially in Central Asia. It was used 
for ornament (a bead in Kayrak-Kumy) and in copper alloying to produce a solid 
alloy. Antimonite deposits for antimony production are found in the Kashka- 
Darya at Shut, on the Zeravshan ridge at the Marguzar lakes, in Fergana on the 
Sokh and Shakhi-Mardan. But the author is unaware of Andronovo objects being 
recovered from there. 

Mining tools are found in the mines, in work stations and in miners' 
settlements. They are uniform along the whole natural habitat (Kuztetsova and 
Teplovodskaya 1994). Wedges were fashioned from rough polished stone; there 
were large stone hammers, weighing up to 40kg, with notches to facilitate 
hafting; there were so-called hoes, pick-axes, hammers, axes used for working, 
and numerous tools for breaking up the ore, along with pestles and mortars; light 
hacks for exploiting placer deposits were made from Siberian deer horn; tools for 
extracting soft rock were made from animal ribs and shovels were fashioned 
from shoulder-blades (Krivtsova-Grakova 1948: 104, fig. 28,4; Chernikov 1970: 
126, tables 24, 1-3, 18,7; Kyzlasov 1965: 165-168, fig. 1-3; Margulan et al. 
1966: 168-169, tables 23, 32, 33, 40, 43, 47, 48; 1979: 238-247, fig. 126, 160,2, 
162-165, 173, 179-182; Vinogradov 1983: 13; Grigor'ev 1988: 47-51; Kuzne- 
tsova 1989b: 119). Pick-axes and picks were made of bronze. 

There were open and closed mines (Fig. 28). Open mines predominated. 
These comprised: type 1) large quarries of round or oval form; type 2) long 
narrow trenches; type 3) several shallow pits. All types are synchronous, the 
different forms of extraction conditioned by the specific character of the ore 
deposit of the mine. 

To the first type belong, for example, the following mines: Vorovskaya 
Yama which measured 30-40m in diameter, 3-5m deep (Zaykov et al. 1995: 
158); Elenovka, which covered an area 46 x 37m, up to 4m deep; Bakr-Uzyak - 
55 x 35m, 3m deep (Chernykh 1972: 77); Uro-Tobe - 32 x 17m, up to 9m deep 
(Chernikov 1948: 20, fig. 7); Airanbai - 45 x 50m; Petro - 16-18 x 8-10m, 2m 
deep; Kresto 2 - 30 x 10m, 4m deep (Margulan et al. 1966: 267); the Sarybulak 
works measured 5-35.7 x 13.9m, 2.5-3m deep; Altyn-Tyube from 17-20 x 9-10m 
to 20 x 20m and 27 x 15m, l-3m deep (Zhauymbaev 1984a: 115, 117, fig. 2). 



MINING, METALLURGY AND THE METAL INDUSTRY 9 1 

Many quarries in central Kazakhstan are much larger, e.g., the Kenkazgan quarry 
was 500-1000 square meters (Alekseev and Kuznetsova 1980: 4-8; Kuznetsova 
and Teplovodskaya 1994). 

To the second type belong the Ushkatta 1 mine with a trench 130m long and 
12-20m wide; Stepnyak, where trenches spread one after another along the vein 
for several hundred meters; the Stalinskiy mine with parallel arrangement of 
several narrow trenches (Chernikov 1948: 15, fig. 3, 6); Sayak with 85 oval pits 
spreading along 1km; Tesik-Tas with 16 oval pits arranged in a chain; Dzhaz- 
kazgan with a 4-5km chain of long trenches (Margulan 1972: 19); the Altyn- 
Tyube extraction site of 9-33 x 2-3m (Zhauymbaev 1984a: 117, fig. 2). 

To the third type of sites belong the Ushkatta 2, 3, 4 mines, Ust'-Kaban, Bes- 
Tyube, Kresto 3, Yashilly Sarybulak, Altyn-Tyube (Chernikov 1948: 21, fig. 8; 
Margulan et al. 1966: 267; Zhauymbaev 1984a: 115-117). 

Closed mines are typified by vertical mine shafts and closed galleries. Type 4 
mines comprise gallery-pits narrowing downwards to a depth of 30m, sometimes 
in steps, into abruptly falling veins; these have been found on cassiterite 
deposits, e.g., in Kara-Goin, Mynchunkur, Alabuga (Chernikov 1970: 121, tables 
20,1,3, 21,1,3). To avoid collapse the ancient miners left arches of native stone 
that divided a gallery into compartments. Type 5 mines comprise closed 
galleries — corridors set on the horizontal or inclined from the hill slope. Small 
galleries with a narrow manhole where it was possible to work only in a 
horizontal position are known from Ushkatta, Kuseevskiy, Dzhazkazgan, Petro, 
Zlatoust, and Altyn-Tyube. Large industrial sites of this type are at Cherdoyak, 
the Karagoyn mines (Chernikov 1949, fig. 2,8; 1970: 122, table 22), Dzhez- 
kazgan, Kyzyl-Espe, Kenkazgan, and Karadzhal (Margulan 1972: 19). Column- 
like pillars of native rock were sometimes left; in Kazanchunkur stone support 
columns are known. In spite of these precautions, galleries did collapse and the 
skeletons of miners are often found in ancient mines (Lev 1934: 21). 

To penetrate the stone Andronovo miners heated the rock and then applied 
water to cool it so that it cracked and was accessible to stone and copper tools 
(Chernikov 1948: 24). The extracted rock was sorted out near the mine. Then on 
the benefication grounds, on the banks of a water source, the ore was crushed 
and washed. Such grounds have been discovered in Elenovka, Ushkatta, 
Stepnyak, Stalinskiy Rudnik, Aul'naya Ploshchad', Bes-Tyube, Dzhezkazgan, 
Kresto, Zlatoust, etc. In central Kazakhstan, where the deposits were situated in a 
desert area, special stone -built dams were constructed to hold flood waters to 
wash the ore (Korgantas, Keregetas, Zhetumshoky, Altynsu, Sorkuduk, 
Taskuduk, Milykuduk, Kipchakpay; Margulan 1979: 263-270). In settlements 
they deposited the washed and crushed ore in pits, 3m in diameter and lm deep, 
and covered them with stone plates (Chelkar, Milykuduk, Atasu; Margulan 1979; 
Kadyrbaev and Kurmankulov 1992). 

Metal smelting took place in the settlements. Remains of smelting have been 
discovered on monuments of all types across the entire Andronovo region. 
Spectacular traces of metallurgical production were already known from the 
early settlements of the Sintashta and Petrovka types. Metallurgical furnaces, 
numerous deposits of slag, drops of copper, and stone metal processing tools are 
found at Sintashta, Arkaim, Kumak, Ust'e, and Kulevchi 3 (Vinogradov 1982: 
97, 98; Grigor'ev 1988: 51, Zdanovich 1989: 184). 



92 CHAPTER SIX 

From the middle of the 2nd millennium BC the main sources of raw mater- 
ials were the native deposits of the Urals and Kazakhstan, and copper was repla- 
ced by bronze. There are numerous traces of slag, copper drops, pieces of ore 
and other evidence of bronze casting on all Elenovka settlements. At Elenovka 
and Tursumbay two-chambered furnaces are found. At Shandasha casting was 
accomplished in a special production room, with slag and a knife mold recovered 
from a furnace. The furnace was connected with narrow grooves, covered with 
plates and filled with ash and slag; into the grooves ran narrow channels into 
which nozzles were inserted. Major industrial complexes (hearths, large pits 
filled with slag and layers of coal, two-chamber furnaces and associated 
fragments of ore, concentrations of slag, copper plates, sometimes crucibles and 
fragments of casting molds) are known from the settlements of Sarafanovo, 
Alekseevka, Burli, Petrovka 2, Novonikol'skoe, Malokrasnoyarka, Novoselovo 7 
(Chebakova 1975: 92, 93; Krivtsova-Grakova 1948: 105, 106; Zdanovich 1988: 
41, 53; Chernikov 1960: 43, 128). 

The most numerous traces of bronze casting production are in central 
Kazakhstan. In the Alakul' settlement at Entuziast 1 a three-chambered bronze 
casting furnace with a stone dome was studied; nearby there were fragments of 
clay casting molds, ore, slag, and metal drops (Tkachev 1987: 31; 1991: 8). 

The Final Bronze Age settlements of Kent, Upais, Buguly, Karkaralinsk, 
Suukbulak, Aynakol', Sorkuduk, and Milykuduk have yielded both production 
complexes and workshops with copper foundry furnaces (pit form, 1-2 to 3-4m 
in diameter, 0.5 to 2m deep, sometimes covered with slabs). There were hearths 
and concentrations of stone mining tools, casting molds, slag, metal plates 
including lead plates up to 5kg in Suukbulak (Varfolomeev 1991: 8,12; Akishev 
et al. 1965: 105; Valukinsky 1948; Margulan 1979: 233; Kuznetsova and 
Teplovodskaya 1994, 1, ch. 2). The appearance of specialized workshops reflects 
the great progress in metal processing in the Final Bronze Age. 

The most spectacular metallurgical complexes have been investigated at the 
settlement of Atasu (Margulan et al. 1966; Kadyrbaev and Kurmankulov 1992; 
Kuznetsova and Teplovodskaya 1994). The whole area of the Atasu settlement is 
peppered with trenches and pits connected with metal processing; there is much 
ore, slag, copper plates and ore crushers on the surface. There are hearths in 
many of the houses. The production complexes are grouped in the center of the 
settlement. The Atasu furnaces were of various constructions, depending on fun- 
ction. For the preliminary heating of the ore there were large clay covered pits, 
2m in diameter, 1.5m deep, with a channel for bellows. The smelting of ore took 
place in furnaces with deepened bottoms. Furnaces for casting metal were pre- 
pared in a pit covered by stone slabs and clay. A smoke flue comprising a chan- 
nel 0.15 — 0.2m deep and from 8 to 12m long led to the furnace. Copper plates 
up to 1.5kg were found near the furnace. Two-chambered furnaces with a cru- 
cible and an air-intake chamber were employed for refining black copper oxid- 
ized ore. They are known from many Andronovo settlements, including the Elen- 
ovka micro-district, as well as in Tuva and in Khakassiya (Sunchugashev 1975). 

The smelted copper flowed down, sometimes into a vessel. At Ushkatta 2, 
Kiimbay, Spasskoe, and Novoselovo there were vessels with copper slag 
attached. The metal solidified in the form of a flat round cake, 10-20cm in 
diameter, and weighing 3 -5kg. Such plates are found at Stepnyak, Sol'-Iletsk, 



MINING, METALLURGY AND THE METAL INDUSTRY 93 

Atasu, at the Kargalinsk and Berdyansk mines, at Borybas, on many Elenovka 
settlements and on sites of the Timber-grave culture (Krivtsova-Grakova 1948: 
106) where they appeared from Andronovo metallurgists (Chernikov 1960: 129; 
Popov 1964: 262; Sal'nikov 1951a: 126; Margulan et al. 1966: 212). 

To improve the quality of the metal tin was artificially added to copper, 
resulting in a bronze alloy with greater solidity. In Central Asia and rarely in 
central Kazakhstan we find alloys of antimony, lead and zinc. Sometimes 
cassiterite, present in the copper ore itself, was deliberately added. The molten 
metal was initially turned into a bar (Chernikov 1960: 130; Bogdanova- 
Berezovskaya and Naumov 1962: 205; Kuznetsova 1987: 44), usually composed 
of pure copper. Metal processing was often accomplished in the same kilns as 
those employed for firing ceramics. 

Objects were produced by forging and casting. Stamping and chasing were 
also used for decoration. Single and bi-valve molds, made of stone and clay, 
were used. In the Final Bronze Age there were undoubtedly bronze molds, e.g., 
the Tomsk mold. Clay molds were used by both the Timber-grave and Samus' 
cultures. In the west Andronovo region and probably in eastern Kazakhstan stone 
molds prevailed; in north and central Kazakhstan clay molds were also 
employed. The latter are found especially often on sites of the Final Bronze Age 
(Sal'nikov 1951a: 128; Zdanovich, S. 1979: 12). Tri-partite molds appeared only 
at the end of the Bronze Age; a plate from the Shamshi hoard was cast in such a 
mold (Ryndina et al. 1980: 168). Each mold was intended for the casting of a 
single object. Matrices for producing several objects, a characteristic of 
Ukrainian metal working, are rare in Andronovo territory, although they are 
known especially in the Final Bronze Age (Kundravinskaya, Aleksandrovskaya, 
Malokrasnoyarka, Alekseevka) (Sal'nikov 1967: 52.15; Chernikov 1960, tables 
38,1,16,1-4; Kuz'mina 1966: tables 3,10; Evdokimov 1975a: fig. 6,8 ). 

Use of the cire -perdue technique was rare. Casting about a core was used for 
the production of socketed and lugged objects. Crucibles in the shape of a footed 
cup are known from Kipel', Elenovka, and Kamyshnoe 1, and clay or stone 
crucibles in the form of cups with vertical or horizontal handles have been 
recovered from Ushkatta 2, Kambulat 1 , Alekseevka, Zamaraevo, mines on the 
Kurchum (Chebakova 1975: 99, fig. 4.3; Krivtsova-Grakova 1948: 106, fig. 32; 
Sal'nikov 1951a: 128; Maksimova 1959: 126, fig. 9). Cast tools required further 
processing with the sharpening of blades and the smoothing of casting seems. 

Andronovo metallurgists worked out the optimum technologies of metal- 
lurgical production, taking into account the function of the object, alloy compo- 
sition, and thermal and mechanical processing regimes (Degtyareva 1975: 12- 
16). The majority of tools was cast in bronze, containing 1-12% tin (averaging 
7%), with further finishing of the whole object (20-40% reduction) and streng- 
thening the working part by hot forging (with 70-80% reduction). Tools that had 
a simple form such as awls, hooks, and chisels were forged from cast blanks. For 
striking tools (picks) and for sickles a softer more plastic metal was needed, and 
so these were made of pure copper or from bronze with a low (less than 2-3%) 
tin presence. For cutting, piercing and chopping objects (celts, axes, adzes, chis- 
els, awls, spears, daggers, knives and arrows) a stronger metal was required, and 
here we find bronzes with 5-7% to 15-25% tin content employed in their produc- 
tion. The blade was annealed for added strength (reduction at 20-40%). Ornam- 



94 



CHAPTER SIX 



ents such as mirrors and plates were produced in single molds; bracelets, ear- 
rings, and beads were forged by cold casting from a cast blank. To provide the 
plasticity necessary in finishing tools and weapons annealing was accomplished 
at 700-800 degrees (Chernikov 1951: 140-161; Chernykh 1970: table 3; Degtya- 
reva 1975: 12-6; Ryndina et al. 1980: 159-170; Korenevsky 1977: 49, table 2). 

Metal processing technology was universal across the entire Andronovo area 
and flourished in the Final Bronze Age, when temperature regimes, processing 
methods and alloying were standardized (Kuz'mina 1966). However, in bronze 
casting there appear local distinctions (Table 4). The concentration of tin 
diminishes according to distance from cassiterite deposits (Chernikov 1948, table 
3; 1960, table 48). 



Region 


Copper 
objects 


Bronze 
objects 


Among them with 
typical tin content 


Typical tin 
content 


E. Kazakhstan 


no 


100 


75 


12-26 


Semirech'e 


few 


88.8 


92 


1-12 


C. Kazakhstan 


29 


62 


78 


6-18 


N. Kazakhstan 


28 


61.4 


74 


1-12 



(After Degtyareva 1985) 
Table 4: Composition of metal industries according to region in % of metal objects 



In the western Andronovo we find tin in only 62% of the objects, and in the 
Volga-Ural chemical group two-thirds of the objects were made with the deliber- 
ate alloying of tin; in the Elenovka-Ushkatta group it was less than a half; it is 
still less in Timber-grave metallurgy where only one third of the cast objects 
were of tin bronze (Chernykh 1970: 16, 21, 22, 1 1 1). Only 35% of the tools were 
tin alloyed and made of metal from the BK group, and one third of the objects 
were of Elenovka-Ushkatta copper (Korenevsky 1977: 49, tables 1, 2). It proves 
that tin was very expensive in the west and it derived from eastern Kazakhstan 
and perhaps partially from central Kazakhstan and Central Asia. 

In bronze casting ore from nearby deposits was mainly used. On settlements 
of the Elenovka microdistrict we find Elenovka ore, at Atasu the ore derives 
from Kenkazgan, at the Kayrak-Kum sites it came from Naukat, etc. It raises the 
issue of several independent local centers of metallurgy and metal processing 
across Andronovo territory (Kuz'mina 1966: 92-94). A center of production is 
indicated by: 1) the presence of copper deposits suitable for exploiting in the 
Bronze Age; 2) traces of exploitation of these deposits by the Andronovans; 3) 
proximity of the composition of the metal objects with the ore from the deposits; 
4) evidence for metallurgy and metal processing on settlements; and 5) the 
spread within a center of a certain limited set of metal types, sometimes 
including specific types typical for a given center. 

The ore from the Elenovka center in the Urals is characterized by the inclu- 
sion of silver, antimony and arsenic (Chernykh 1970: 22, fig. 9, 21); the ore in 
northern Kazakhstan contains gold but no arsenic, no or little antimony, lead and 
nickel; central Kazakhstan ore contains admixtures of lead, iron, manganese, also 
of zinc, silver, cobalt and nickel (Kuznetsova 1989: 100); the eastern Kazakhstan 
center is characterized by its high content of lead and especially antimony, it has 
no gold and nickel (Chernikov 1951: 142,150, tables 1,2); the Semirech'e center 
is marked by its high content of nickel, lead, arsenic and also bismuth, silver and 



MINING, METALLURGY AND THE METAL INDUSTRY 95 

antimony (Kopylov 1955; Kuz'mina 1966: 109, 110; Ryndina et al. 1980: 154, 
fig. 1); Fergana is characterized by the presence of zinc along with its high con- 
tent of antimony, lead, arsenic, silver and bismuth (Kuz'mina 1966: 107; Ruza- 
nov 1982) and finally, the Andronovo objects from the center of Central Asia 
contain much lead, antimony and arsenic (Ryndina et al. 1980: 157). One will 
probably be able to identify metallurgical centers in Siberia, especially in the 
Altai, in the future. Each center is characterized by a specific assemblage of 
metal objects (Chernikov 1960; Kuz'mina 1966), as well as preferred production 
traditions (Degtyareva 1985: 18-22). In every center proposed the bronze objects 
reveal a chemical composition involving the presence or absence of elements 
corresponding to the ore from local deposits. However, even at sites near mines 
there are objects made from imported metal. Thus, in the Elenovka micro-district 
alongside the dominant Elenovka copper there is also metal of the VK and VU 
groups (Chernykh 1970, table 3, #1433, 1435, 1437-9, 1445, 1449, 1452). A still 
more varied metal composition is observed on sites situated far from mines. In 
the Tasty-Butak cemetery there are objects of three types of alloys and an object 
regarded as an Altai import (Bogdanova-Berezovskaya and Naumov 1962: 203- 
206, table 6). One object from the Shamshi hoard was imported (Ryndina et al. 
1980: 154). 

Thus, in contrast to pottery metallurgy was not a domestic craft among each 
community of Andronovans. Metallurgical production was a special craft that 
was not only intended to meet the demands of the local community but was also 
produced for exchange (Chernykh 1970: 112, fig. 30, tables 1-5). Metal from the 
Kargaly mine was distributed as far as the Danube (Chernykh 1996: 71). Metal 
from Elenovka-Ushkatta (Map 6) was utilized by related Andronovo tribes 
(Khabarnoe, Uvak Mechet-say, Alakul', Chernyaki 2, 3, Kipel', Novo-Burino), 
mixed Timber-grave-Andronovo tribes (Gerasimovka; Fig. 59), and the bearers 
of other cultures: Pre-Kazan' (VII Lebedinskaya), Pozdnyakovo (Borisoglebskiy 
cemetery) and especially the Timber-grave tribes, where objects of Elenovka 
metal were spread from Bashkiria (Beregovskoe settlement) to Kalmykia 
(Elista), and the Don (Ilmen, Mazurka). They were especially numerous on sites 
on the Volga (Yablonovka, Politotdel'skoe, Atkarsk, Berezhnovka, Rovnoe, 
Molchanovka, Potemkino, Pokrovskiy, Skatovka, Karamysh, Chardum). N. L. 
Chlenova (1983a: 56) composed a map of the distribution of ore and gold objects 
(table 8, map 5) on the basis of our data and the analyses of E. N. Chernykh. In 
Timber-grave metallurgy one can discern a small group of tinned bronzes from 
the Altyn-Tyube deposit near Karaganda (Chernykh 1970: 17). Bronze objects of 
Andronovoid types are also numerous in the northern Andronovo cultures. 

There are no data supporting the exchange in bar-form or ready-made 
objects. Probably both of them took place. The few bars and traces of metal 
processing on Timber-grave settlements on the Volga, and the typological 
closeness of Timber-grave objects, especially ornaments, with those of the 
western Andronovo suggest that the trade was in the form of finished objects 
with the Timber-grave people. In contrast, the majority of Andronovo 
settlements, even those far from deposits, e.g., the settlements of Kipel', Novo- 
Burino, Bakhtinskoe, Alekseevka (Sal'nikov 1951a: 127, 128; 1967: 337; 
Krivtsova-Grakova 1948: 104-107), are characterized by molds, drops of copper, 
slag and even pieces of ore, although not as numerous as in the micro-districts 



96 CHAPTER SIX 

surrounding a mine, but proving that metal processing was done by every 
community to satisfy its demands. 

The Andronovans outstripped the neighboring Eurasian tribes in mining, 
metallurgy and metal production, supplying bars and ready-made objects. 

Quantifying the amount of metal extracted by the Andronovans is extremely 
difficult as in ancient times a large amount of native copper and gold was used 
which is impossible to take into account; also not all deposits exploited by the 
early populations are known nor is the scale of mining in the Iron Age known. E. 
N. Chernykh (1996: 71) calculates that during the Eneolithic and Bronze Age the 
giant Kargaly mine yielded 1.5-2 million tons of ore. Extraction of ore in Vorov- 
skaya Yama reached 6,000 tons of ore, i.e., 10 tons of copper (Zaykov et al. 
1995: 161). At the Kenkazgan deposit on the basis of the Atasu settlement 
extraction is estimated at 800,000 tons of ore which would yield 30,000-50,000 
tons of cast copper (Alekseev and Kuznetsova 1980: 4-8). More reliable are 
calculations for cassiterite extraction in eastern Kazakhstan because it was 
exploited only in the Bronze Age. S. S. Chernikov (1960: 135) estimates that 130 
tons of tin ore was extracted which constitutes 130-150kg of annual tin extrac- 
tion, reckoning on about 500 years of exploitation. 

The organization of labor in Andronovan metallurgy is problematic. S. S. 
Chernikov (1960: 132-136) considered that deposits were exploited only in the 
summer by small clan groups of 8-10 people; moreover, child labor was used in 
the narrow pit-faces. However, at the large deposits of Elenovka and in central 
Kazakhstan, where there is evidence of settlements of whole groups of metallur- 
gists, the scale must have been considerably larger. 

G. B. Zdanovich (1989) believes that the Sintashta population, living near the 
Ural mines in a fortified settlement of the Arkaim type, had achieved the stage of 
urban civilization which is characterized by craft specialization. The success of 
these tribes in metal production and in military affairs secured their supremacy 
of the Eurasian steppes and preconditioned their society for stratification. How- 
ever, one should not overestimate the level of their culture. There are no traces of 
social and property differentiation within the settlements, there is no evidence of 
communal -scale craft specialization. The specific character of their cattle-raising 
economy and the vastness of the steppe led to the extensive way of their culture 
development. Progress in metallurgy and in metal production was not based on 
the increase of productivity by the growth of craft-specialization and division of 
labor into a special branch of society, but rather by the introduction of new tech- 
nologies: copper was replaced by a bronze alloy, the development of alloying 
methods to suit the function of the implements, the creation of optimum thermal 
regimes and forging methods for ready-made casting. All these innovations 
spread quickly across the whole Andronovo region and technological progress 
was realized through standardization. 

E. F. Kuznetsova (1987: 44) and A. D. Degtyareva (1985: 200) suggest that 
metal processing in the 15th 13th centuries BC should be characterized as an 
independent craft, the rise of craftsmen clans in different territories (divided into 
miners, metallurgists-founders, and smiths); this is not confirmed by evidence 
from Andronovo settlements. 

Considerable progress in metallurgical production was achieved in the Final 
Bronze Age. At that time Andronovo tribes started using not only oxidized but 



MINING, METALLURGY AND THE METAL INDUSTRY 97 

also sulphide ores, which demanded additional processing stages: slag and matte, 
production of black copper with the required increase in temperature. It deman- 
ded improvement of the type of metallurgical furnaces, such as that discovered at 
Atasu. Considerable progress in metal processing is proved by new types of 
alloys (copper+lead+tin, copper+antimony+tin, copper+arsenic+tin) (Kuznetso- 
va 1987; 1994), tripartite molds, new types of weapons and tools with higher 
efficiency. All these innovations presupposed the development of metal 
processing seen in the Early Iron Age: most types of metal tools, daggers, and 
arrows were prototypes of objects used by the Saka and Scythians. They also 
inherited many technological processes and methods, and the introduction of iron 
itself was a logical result of Final Bronze Age progress; experiments in their 
integration are found on many settlements of the 1 3th 9th centuries BC. 

Considerable changes in the steppes during this period are shown by the 
appearance of hoards of bronze objects. There are hoards of two types: family 
and founders hoards (Kuz'mina 1966: 98). Family hoards (Brichmulla, Turksib, 
Sadovoe, Sukuluk, Issyk-Kul', Shamshi, Tuyuk, etc., Figs. 43, 114) contain 
various types of used objects, which were family property. The appearance of 
such hoards reflects the process of property stratification of the late Andronovo 
tribes. The concealment of hoards in the earth indicates the tense situation in the 
steppe, more frequent military confrontations, which is proved by the spread of 
numerous types of new defensive weapons and the appearance of cheek-pieces 
that were used by mounted warriors. All this is evidence of a uniform process 
connected with the transition to nomadic cattle-breeding. 

The second type is the founder's hoard. They contain metal, molds, blanks 
and a single type of object, sometimes cast using a single matrix. Sukuluk 2 with 
its sickles is an example of a founder's hoard. The existence of such hoards 
points to metal processing involving a separate branch of craftsmen and its clan 
organization. However, in contrast to the Danube and north Pontic region where 
dozens of founder's hoards are found, the discovery of such hoards over the 
whole Andronovo territory is rare. Consequently, metallurgical production 
among the late Andronovo tribes remained mainly a communal affair, although it 
was highly developed and involved production for export. The character of 
settlements near mines shows that the extraction of ore and its processing was 
done by people possessing equal rights, living in non-fortified settlements in 
large houses, which did not differ in wealth and which belonged to large family 
communities. There are no data to support the existence of labor divisions 
between miners, metallurgists, smiths and those who fashioned ornaments. 
Judging by the numerous traces of metal processing, found in every settlement, 
in every house in the micro-districts near mines, every large family was engaged 
in all the operations associated with the production of metal. 

On every settlement of metallurgists animal bones were found proving that 
cattle-raising was the leading occupation of each family; utensils of local 
production and traces of other domestic activities also prove that metallurgy and 
metal processing were not at the stage of a specialized craft. 

The organization of production among the Andronovans differed consid- 
erably from what is known for other states of the Old World. Production techno- 
logy was different: in the Caucasus, in the Near East, Iran and south Central Asia 
the lost -wax technique was employed. It demanded great effort from professional 



98 CHAPTER SIX 

craftsman, but resulted in the production of unique objects made to individual 
order. Casting in reusable composite cast molds dominated in the Eurasian 
steppe. This production required neither great labor effort nor high professional 
skills and made it possible to produce objects for export on a large scale. 

The methods of organizing mining production were also different. In King 
Solomon's mines on Timna mountain (Rothenberg 1962: 9-43) during the 10th 
century BC, extractions were made to meet the demands of the Temple of 
Jerusalem. Several open mines, 50-300m long and 10-30m deep, with horizontal 
galleries were discovered there. Stone mining tools were analogous in type to the 
Andronovo (Rothenberg 1962: 1,12,14); processing sites near mines and hearths 
for ore smelting and casting in metallurgists' settlements several kilometers from 
the mines are also analogous to Andronovo. But the settlements were quite 
different. The houses were extremely poor: they were small or subdivided into 
small rooms by partitions; the area of the two main settlements was surrounded 
by defensive walls with towers for supervising the slave-miners. In the Sinai 
mines there were numerous interconnected galleries on different levels in which 
hundreds of slaves worked (Garland and Bannister 1927). Such a scale of 
manufacture was required for the excessive demands of the nobility and priests 
and was possible only in a state in which slaves worked on the Pharaoh's order. 
In other Near Eastern societies metallurgy was the occupation of special clans or 
of craftsmen isolated from the community but subordinated to the temple or 
sovereign. In the 3rd-2nd millennium BC there was specialization among them, 
and one could find miners / ore diggers, copper founders, smiths, and ornament- 
makers (Chernykh 1972: 183-4, 192). In the time of Hammurabi there was the 
nappahu - preparatory smith ('blowing fire in the hearth'), gur-gurru - copper 
founder, nappah-hurasi - copper smith. These linguistic data are in good accord 
with archaeological data: metallographic analysis of objects from the Anau 
culture of south Turkmenia, belonging to the periphery of the Near Eastern 
farming cultures, showed that already in the 3rd millennium BC the metal 
processing of the Anau people was a specialized craft, with a narrow division of 
labor (Terekhova 1975: 41, table 3). In Mesopotamia the narrow professional 
specialization of metal workers-craftsmen occurred even earlier (Maryon 1949). 

The Indo-Iranians knew metal well. They knew copper (bronze), gold and 
silver. Indo-European names for metals relate back to color names with a further 
shift of meaning (Gamkrelidze and Ivanov 1984: 710-714). But there are no 
common Indo-European terms connected with metal processing. Indo-Iranians 
knew the awl, needle, some types of knives, including a knife for slaughtering 
animals, arrow, spear, hammer, and axe. The Indo-Aryans also knew Indra's 
weapon, the vajra-. Ornaments comprised gold earrings, bracelets, pendants, 
plates with holes or small loops for sewing on clothes (Rau 1983). 

The Indo-Aryans were familiar with metal casting: "when Trita is blowing it 
(fire - E.K.) in heaven like the blower (of bellows), he sharpens (the flame) as 
the fire blazing from the blast of the blower" {Rigveda 5.9.5). But in Vedic there 
is no term to denote craftsman. In describing the creation of something, the verb 
At- 'make' is used, in which the act of creation is viewed as a ritual (Elizarenko- 
va and Toporov 1995: 500). The main 'maker' is the god Visvakarman, and the 
main creator and inventor in Vedic mythology is the god Tvastar, who creates 
chariots and weapons {Rigveda 1.13.10; 5.42.13; Satapatha-Brahmana 2,4,3). 



MINING, METALLURGY AND THE METAL INDUSTRY 99 

As far as the Iranians are concerned, in the Avesta there is no such term as 
'smith' and there are no data about a specialized craft (Geiger 1882: 388, 479, 
480; D'yakonov M. 1961: 60f, 362f). The word 'craftsman' (huti) is mentioned 
once (Yasna 19.17). It proves that the crafts, including metal processing, of the 
Indo-Aryans, remained an unspecialized occupation in their homeland. It is 
interesting to note in this connection that in the Sintashta-type cemeteries of 
Kamenny Ambar (Fig. 60), Solntse 2 and Bestamak chariot warriors were buried 
with adzes for chariot making, ore, nozzles and abrasives for weapon production, 
combining warrior and smith functions as well as smith and carpenter, as Tvastar 
and Visvakarman. The Ossetic god Kurdalagon, creator of the wheel and 
weapons of the Narts, corresponds with them in function. 

Mythology concerning smiths is widespread among Indo-European peoples, 
moreover, smith-gods occupy an important place in the pantheon, equal to the 
other high gods, which reflects the equal position of metallurgists in the actual 
society of ancient Indo-Europeans. These linguistic and mythological data 
correspond to the Andronovo and not the Near Eastern character of metallurgical 
production. Arriving in an area of developed craft production the Indo-Iranians 
borrowed special craft terms from local craftsmen. In the Avesta the word 
tanura- 'oven' is borrowed from Sumerian through Akkadian (Herzfeld 1941). 
The word for an oven in the Rigveda was borrowed from aboriginal Dra vidian or 
Munda (Kuiper 1991: 14, 21) as many other cultural terms. 

Studying the question of metal processing in the context of the problem of 
Indo-Iranian origins it is necessary to mention Aryan loans in the Finno-Ugrian 
languages that were spoken in the forest zone of the Urals. These linguistic data 
correlate with archaeological facts about the spread of Andronovo types of 
bronze objects into the taiga and about the creation of Andronovoid cultures 
there, pointing to a steppe homeland for the Indo-Iranians. 

Another direction of Andronovo metal diffusion was to the south, into 
Central Asia and Afghanistan, probably reflecting not only cultural ties but also 
the gradual migration of the Andronovans to the south. 

Thus, the analysis of mining, metallurgy, and metal processing does not 
support a hypothesis about population migration from the Near East in the 
second half of the 2nd millennium BC and it does not contradict attributing the 
Indo-Iranians to the Andronovo culture. 



CHAPTER SEVEN 
TEXTILES AND DRESS 

Andronovans employed felt, fur, leather and wool in the manufacture of their 
clothes. The use of fur is indicated by the recovery of bones of wolf, fox, ferret, 
beaver, and hare on Andronovo settlements. 

Leather production skills is testified by blades and scraping implements made 
from the jawbones of horse and cow found at Alekseevka, Sadchikovo, Shan- 
dasha, Tursumbay, Chaglinka, Yavlenka, Atasu, Kanay, Ust'-Narym, Malo- 
krasnoyarka, Tasty-Butak, Kipel' and other sites. According to G. F. Korobkova 
these objects were used for manufacturing leather (Sharafutdinova 1982: 136; 
Leskov 1970: 39). Andronovans used leather for the sewing of outer clothing, 
caps and footwear. 

Wool was the main material for clothes. It was spun with wooden spindles 
with clay spinners as shown by their discovery in settlements, and the twisted 
threads themselves are preserved inside beads in burials. The simplest way of 
producing woolen things was knitting. In western Siberia in the Andronovo, 
Orak, Pristan', and Ust'-Erba burials, knitted textiles were found (Tugarinov 
1926: 158; Sosnovsky 1934: 95-96; Kiselev 1949: 44, 48; Komarova 1961: 51; 
Maksimenkov 1978: 72). At Orak narrow strips were knit in the Tambur chain- 
stitch manner from rough twisted wool 29 microns wide. They were then sewn 
on in spirals to form a conical cap with a herring-bone pattern (Sosnovsky 1934: 
93-95). At Andronovo, a conical cap was also sewn in spirals from narrow strips, 
not knit but twisted, from thick wool thread. In two other graves at Orak there 
were found narrow strips of fabric fragments from clothes and a cap, plaited in 
galloon weave from untwisted wool using a shuttle, the thread width being 18 
and 22 microns (Sosnovsky 1934: 93-95) 

Weaving also existed. At the Elovka burial cloth of 'diagonal' type woven 
from thin wool threads was preserved (Matyushchenko 1973a: 59). There are 
imprints of cloth from woven linen, made of threads 1.3-1. 6mm wide, on 
Petrovka and Alakul' pots. Such fabric is known from imprints on ceramics and 
a knife from Seyma (Bader 1970: 123). Analogous fabrics of linen and diagonal 
weave were found in Pazyryk (Rudenko 1953: 245, table 25.4). Apparently they 
wove on a primitive vertical loom without shafts, using knitting needles for 
stretching the base and sometimes clay or stone weights. The discovery of 
knitting needles, weights, and a shuttle are known from Andronovo settlements 
(Sal'nikov 1951a: 139); a shuttle was found at Chaglinka (Orazbaev 1970: 134). 

The spinning and weaving of the Andronovo culture are analogous to west 
European techniques that are especially well studied in Denmark (Glob 1947) 
and with east European textiles of the Bronze and Early Iron ages (Pislariy 1981; 
Gavrilyuk 1989: 84-91). Andronovo and Timber-grave spinners belong to a 
widespread ancient type preserved by the Ossetes and Iranian-speaking peoples 



102 CHAPTER SEVEN 

of the Pamirs (Karmysheva 1979: 250-269). A. A. Semenov and G. F. 
Korobkova (1983: 130-132) have shown that twisting and knitting preceded 
weaving. The proto-type loom comprised a wooden frame with a stretched base 
and then the weft was threaded with a needle or kochedyk. The base might also 
be tightened with wooden pegs that had been driven in. A primitive loom 
without weights is preserved among the Mountain Tadzhiks (Pisarchik 1958: 
372, fig. 80-81). 

Terms connected with spinning and weaving derive from the most ancient 
layer of Indo-European heritage. Semantic bundles of words are built from the 
stems *ten and *tek ('draw', 'twist', 'braid', 'spin', 'weave', 'thread', 'web', 
'stripe', 'cloth'; Abaev 1949: 54; 1979 III: 220, 221, 302, 336, 337; Elizarenkova 
and Toporov 1995: 522). The names for wool and weaving share a common 
origin with the word for 'sheep' (Gertsenberg 1972: 56, 57, 228; Gamkrelidze 
and Ivanov 1984: 583, 704-705). In the Rigveda clothes from fur and skin are 
mentioned (Elizarenkova and Toporov 1995: 521). Apparently alongside wool, 
vegetable fibers, hemp in particular, were used (Gryaznov 1956a: 40; IK SSR: 
123). Imprints of a textile hurdle from organic fibers are found on Eneolithic 
ceramics, for example, at Botai. In the Urals hemp processing survived till the 
19th century (Sal'nikov 1961a: 139), and in the Ukraine and Ossetia it is cultiva- 
ted to the present. Cloth made from hemp was found at Pazyryk (Rudenko 1953: 
104). According to Herodotus (4,74) hemp grows in the country of the Scythians, 
where it "much surpasses flax. It grows both by itself and is tilled" (Dovatur et 
al. 1982: 127). In the Indo-European languages names for flax and hemp are of 
common origin: the same word is used to denote home-spun cotton fabric 
(Abaev 1958 I: 513; Gertsenberg 1972: 181, 183; Steblin-Kamensky 1982: 63f). 

Spinning and weaving are very frequently used in Vedic literature. Terms and 
notions connected with them have been studied by W. Rau (1971); some data are 
found in general works on the history of crafts (Rau and Chakrabarti 1975) and 
Indian costume (Mili Chandra 1972; Parpola 1985). Spinning and weaving were 
female occupations; they used primitive looms without spindles. It is described 
in the Rigveda (10.130): abase, on which a weft is used for weaving, is tightened 
on pegs. Vegetable fibers and woolen yarn were employed. According to the 
Avesta, Yima taught people to use the loom. The goddess Anahita, who appears 
as a weaver, is the protectress of female occupations. In Indo-Iranian tradition 
weaving is a ritualized process with cosmological associations. 

The style of women's clothes in the Andronovo culture is reconstructed on 
the basis of the position of bronze and paste beads sewn on the sleeve cuffs, 
collar and hem. They are known from Alakul', Petropavlovsk, and Ataken-say 
(Sal'nikov 1951a: 140; Kuz'mina 1986b: 978). The clothes comprised a long, 
straight dress, over knee length, with long rather wide sleeves reaching the wrists 
and with a rounded neck. The front of the dress is often decorated with bronze 
plates. The dress was tied up with a waistband with attached amulets from 
perforated animal teeth. It has been claimed that in Orak and Petropavlovsk 
clothes were dyed red with organic dyes (Kiselev 1949: 48; Sal'nikov 1981a: 
140). Red color played a great role in Aryan ideology (Elizarenkova 1995: 481f). 

The costume was accompanied by ornaments. A standard set included a pair 
of earrings, or temple rings, one or two bracelets sometimes with spiral ends, a 
string of beads on the ankles and some breast plates, often of perforated shell. In 



TEXTILES AND DRESS 103 

some graves of Sintashta-Petrovka, Alakul' and mixed types, women were 
buried in rich ceremonial dress. This consisted of a hoop-shaped diadem or 
pendant, several bracelets, spiral finger-rings, plaited ornaments from several 
layers of connected beads and belts with beads, spacers and attached plates and 
pendants including spectacle and cross-like shapes (Usmanova and Tkachev 
1993; Usmanova and Logvin 1998; Vinogradov 1998). 

Men apparently wore unfastened double-breasted clothes of caftan type and 
trousers (Sosnovsky 1934: 95-96; Kiselev 1949: 48; IK SSR: 123; Kuz'mina 
1986b: 98, 99). 

Andronovans wore boots. In burials at Orak and Pristan' in Siberia leather 
heelless boots with high tops, sewn with threads of tendon, tied round by a lace 
with stringed bronze beads, were preserved (Komarova 1961: 35, 50; 
Maksimenkov 1978: 14). According to numerous findings of such strings of 
beads on the legs of the buried, such footwear was used both by men and women 
across the whole Andronovo area. N. A. Avanesova's supposition (1981: 36) that 
women wore wide trousers has not been proved. 

The headgear of Andronovans were found in female and male burials of 
western Siberia at Andronovo, Orak and Pristan' (Sosnovsky 1934: 95-96; 
Kiselev 1949: 48; Komarova 1961: 43, 44, 51, fig. 3; Maksimenkov 1978: 14, 
72). They wore knitted caps with a high conical top and separately attached 
earflaps. At Orak the remains of a leather cap with a pointed top were also found 
(Sosnovsky 1934: 95-96; Komarova 1961: 51). A reconstruction of Andronovo 
headgear was made by M. M. Gerasimov (Krivtsova-Grakova 1948, fig. 8.63). 
Ceremonial headgear was decorated with beads and plates at Alekseevka, Tasty- 
Butak, Ulyubay (Krivtsova-Grakova 1948: 66; Sorokin 1962b: 63; Boiko 
Tatarintsev 1976: 72, Vinogradov 1994: 195). 

The complex of Andronovo clothes was perfectly adjusted to the nature and 
life of Eurasian shepherds and has analogies and prototypes among the steppe 
cultures. The earliest evidence of this type of dress comes from certain Pit-grave 
culture burials in the Ukraine, namely, leather boots 0.3m high from Gradeshka 
(AO 1983: 360) and a felt cap with leather application from Aleksandrovka (AO 
1979: 301). A cap was reconstructed from a rich Timber-grave female burial 
from the Zolotaya Niva cemetery on the Volga (AO 1976: 150). It is embroi- 
dered with bronze spacers, pendants and plates with a Mycenaean-type spiral. 

The types of clothing from the Eurasian steppe of the Bronze Age developed 
over three millennia. The dress of Iranian-speaking Saka and Scythians is easily 
reconstructed on the basis of descriptions provided by Greek authors (Herodotus 
6.87,88; 7.64; Dovatur et al. 1982: 392-394), representations by Greek artists 
and toreutics such as the Chertomlyk vase, the rhyton from Karagadeuakhsh, on 
vessels from Kul'-Oba, Merdzhany, Gaimanova, the Voronezh kurgan, on a 
pectoral from Tolstaya Mogila, on the comb from Solokha, plates from Kul'-Oba 
and other sites (Stepanov 1916: tables v, vi; Artamonov 1966: tables 147, 148, 
153-155, 166-7, 174, 195-8, 201, 203, 224, 226-29, 253, 255, 318) and numerous 
archaeological discoveries from the Ukraine to the Altai, particularly at Issyk in 
Kazakhstan (Akishev 1978, fig. 62, 63; K. Akishev, A. Akishev 1980), at 
Pazyryk (Rudenko 1953: 111,112) and Ak-Alakha (Polos'mak 1994: 3). 

Women wore long dresses, boots and pointed caps or a kalathos of Greek 
origin with a long veil (Miroshina 1977; Klochko 1979). Across the whole area 



104 CHAPTER SEVEN 

men's clothes included trousers, unfastened jacket, boots (Klochko 1979; 1984; 
1992) and a pointed cap. Herodotus (7.64) writes that the Saka "wore pointed top 
caps, standing straight". In the nomadic environment the cap was an ethnic 
attribute, and ceremonial headgear was used as the insignia of a king (Kuz'mina 
1958: 124-125; 1977: 91, 92; 1981b: 46-49; Akishev 1978: 43-44; Akishev, K. 
and A. 1980: 14-31). These clothes are close to those of other Indo-European 
peoples of Eastern Europe; the Thracians, Phrygians and Hittites all wore caps. 

The complex of clothes established in the steppes of the Eneolithic and 
Bronze Age and inherited by the Saka-Scythians has survived among the modern 
Ossetes (Kaloev 1971: 174-184) and Pamirs (Andreev 1958: 243-246, 416, fig. 
50, 90.1), and partially among the Kurds. From the Indo-Iranians and Iranians 
unfastened clothes spread to their northern Finno-Ugrian neighbors, the Khanty 
(Prytkova 1953: 123-233; Moshinskaya 1978), and trousers to Chinese horse- 
men. The dress of the earliest Eurasian pastoralists greatly influenced the 
costume of Turkish steppe peoples, who replaced the Saka in Kazakhstan and in 
the steppes of Central Asia. Not only did they adopt from their predecessors the 
wearing of unfastened clothes, trousers, and boots, but in many cases they 
adopted the word for these items also, most significantly the word for the caps to 
which most male and female headgear are referred to (Sukhareva 1954; 1982; 
Zakharova and Khodzhaeva 1964). 

This complex of clothing, ecologically conditioned by nature and the cattle- 
raising way of life in the steppes would be climatically unreasonable for farming 
cultures of the Near East and has neither analogy nor source there. According to 
imprints on seals and in toreutics of pre-Achaemenid Iran, Bactria and Margiana 
the men wore long, wide clothes or hip bands of a skirt -type and women wore 
dresses with a wide skirt, had no headgear, were barefooted or wore sandals, and 
had a short haircut (7000 ans..., tables 3, 19, 30: Houston 1954, Ligabue and 
Salvatori 136, 137, 163: 106-113; Sarianidi 1998: 11, 18, 39). Native Near Eas- 
tern peoples preserved variants of these clothes in the 1st millennium BC, which 
are illustrated, e.g., in Persepolis (Dutz 1971). The clothing of the Iranian- 
speaking peoples of Iran, Afghanistan and south Central Asia of the Achaemenid 
period differs sharply from Near Eastern dress and is similar to that of the Saka- 
Scythians. 

The clothing of the Chorasmians, Sogdians, Bactrians as well as Persians and 
Medes themselves are well known from reliefs at Naqsh-i-Rustem, Persepolis, 
on objects of toreutics of the Oxus treasure, from Greek and Persian seals, and 
from the Pompei mosaic with a battle scene between Alexander and Darius, etc. 
(Sarre and Herzfeld 1910: fig. 37, 39, 54, 256; Dutz 1971: tables 7, 11, 13, 15, 
17; Dalton 1964, fig. 49, tables 4, 10, 14, 15, 12; Boardman 1970: 882). The 
complex of clothes depicted differed only in certain details of fashion according 
to different tribes such as distinctions in caftan (candiz), trousers, boots and cap 
(Thompson 1965; Beck 1972). H. Bailey (1955: 7-12) demonstrated a common 
Indo-Iranian origins for the names of parts of clothes. Sanskrit kurpasa, Iranian 
kurtak, Russian kurtka derive from the word kur 'neck', 'throat'. In some 
languages it denotes a shirt or jacket; in others a cap, for which there is a term 
kulah. Of Indo-Iranian origin are also names for shirt, trousers, breeches. These 
words were borrowed by the Chinese. Another name for trousers (shalwar) is 
sharovar (wide trousers), which is of ancient Iranian origin, as well as the name 



TEXTILES AND DRESS 105 

of the burkha, the Caucasian felt cloak and woman's head scarf and veil (Abaev 
1949: 53; 1979 III: 26, 27, 79, 125). Words for personal ornaments also have an 
ancient origin: 'necklace', 'neck-band' are Indo-European while 'ornament' and 
'belt' are Indo-Iranian words (Gertsenberg 1972: 25, 189, 191, 195). 

There are not enough data for the reconstruction of the clothes of the Vedic 
Aryans. They wore outer and under clothing with names of Indo-European 
origin, such as the belt and gold decorations: earrings, bracelet, ring, neck 
(pendant) and ear (earrings); gold plates for the decoration of the breast, and, 
what is especially important, they wore plaits (Elizarenkova 1989: 449; 
Elizarenkova and Toporov 1995: 52; Parpola 1985). According to the Avesta, 
Zoroastrians wore a shirt, footwear and a belt as well as a cap made of sheep 
leather (Dhalla 1922: 174). Anahita and Vayu had boots, and Vayu's were 
golden (Yasht 5.64,78; 15.57). According to the ArdvTsur Yast (5.127) Anahita 
was adorned with a pendant and golden temple rings (but according to some 
scholars this is a later text describing the statue of the goddess, erected by 
Artaxerxes). Anahita wears a beaver fur coat and a shirt girded by a richly 
decorated waistband (Yasht 1.17; Yasna 9.26). A twisted belt was a sign of 
membership of a social group among all Indo-Iranians and it was fastened in a 
special way during a dedication rite (Manusmrti II: 42, 43). 

The attribute of an Indo-Iranian priest was his staff. It was handed to Yima 
by Ahura-Mazdah. In modern Ossetia the staff is a symbol of spiritual power of a 
venerable old man, performing priestly functions in the sanctuary of Rekom 
(field notes of the author 1974 ). 

Although the hair styling of the Andronovo male is not known, the female 
hair styling is reliably reconstructed from the position of decorations. Androno- 
van women had two plaits reaching below the waist. They plaited strings of 
beads with hanging oblong bronze plates (Usmanova and Tkachev 1998; Vino- 
gradov 1998). Such ornament was already established in burials of the Early 
Bronze Age near Azov and at Chapaevka, in the Saratov region (AO 1981: 109; 
1983: 159). The same hair style — two plaits with decoration twisted with laces 
and beads — is still worn by women in the Pamirs (Andreev 1958: 252, 414, fig. 
89.3-4) and in some districts of India. According to gliptic materials, in the 
Achaemenid period across the whole of the Near East Iranian peoples wore plaits 
hanging behind their heads, namely Persian and Bactrian women (Dalton 1964: 
103; Boardman 1970: 879, 880, 891, 892, 903, 964, 990; Kuz'mina 1979: 35-40, 
fig. 1, 2). Thus, headgear, costume and footwear were most important ethnogra- 
phic indicators, sharply separating the Iranian peoples of Iran, Afghanistan and 
Central Asia from other peoples of the Near East for whom these clothes were 
entirely alien. But at the same time this dress is analogous to clothes and hair 
styles of Iranian nomads of the 1st millennium BC, and survivals of such a 
complex are preserved by the peoples of the Pamirs and Ossetia. 

The dress of the Indian peoples is extremely various and reflects five 
millennia of the history of the subcontinent. Indian costume is highly 
meaningful. It characterizes regional, ethnic, caste, confessional, social, family- 
status, sex and age differences. Within each group there exist peculiarities of 
clothes, everyday, ceremonial and ritual. This conditions the extreme variety of 
Indian clothing types (Flynn 1971). It is more surprising that in modern India 
there remain strong ancient traditions regarding clothing (Maretina 1977; 



106 CHAPTER SEVEN 

Bulanova 1989; Parpola 1985). The non-Indo-Aryan peoples of southern India 
are descendants of a native population and are also representatives of the lowest 
castes of the whole subcontinent. They wear drapable clothes; males: a hip band 
wrapped over a skirt and dhoti, and females: a sari (Flynn 1971; Bulanova 
1989). This complex of clothes goes back to the pre-Vedic epoch of the 
Harappan culture. 

In the north-west of Hindustan, different ethno-caste groups tracing their 
genealogy to Vedic Aryans preserve in the main another type of dress. Its 
elements bear ancient Indo-Iranian names. Men wear trousers and wide trousers 
(shalvar), shirt or unfastened jacket (kurta) and an outer unfolded caftan; women 
wear a long straight dress (kamiz) (Maretina 1977). This dress finds analogies 
among the Iranian peoples of the Pamirs, the nomads of Iran and the Ossetes, 
and goes back to the Saka-Scythians and finally to Andronovo dress and has 
neither analogies nor prototypes in the Eneolithic or Bronze Age cultures of 
India and the Near East. These clothes are a social symbol. They are worn in 
central and southern India by a group attributing itself to the ksatriya caste. Their 
dress necessarily includes trousers, unfastened caftan and a belt — the priestly 
cord of the twice-born. The ceremonial ritual dress of the Maharaja of Maysur 
who traces his origin to the legendary Rama (field material gathered by the 
author in 1984) goes back to the Indo-Iranian period. The Indo-European color 
symbolism is also preserved: white colored clothes are symbolic of the brahman 
caste, and red - of the maharaja and ksatriya. Decorations have complex ancient 
semantics, e.g., the symbols of married women include the thali pendant and 
mangalsutram necklace. 

The Meo ethnic group originating from the Saka and intrusive into India in 
the 2nd century BC represents Eurasian nomads with a complex of clothes 
depicted in the sculptured portraits of the Saka and Kushan kings in Mathura 
(Pugachenkova 1979, tables 236-237; Vogel 1930; Rosenfield 1967). 

Taking into account the ethnographic importance of costume, and the fact 
that according to linguistic data some of its parts bear Indo-Iranian names, it 
follows that it was formed in the homeland. This dress has no source in the Near 
East, and its genesis is found in the Andronovo culture. These data serve as a 
serious argument in favor of the Indo-Iranians having a steppe homeland and 
assigning an Indo-Iranian ethnic identity of the Andronovans. 



CHAPTER EIGHT 

TRANSPORT 

Wheeled transport was widespread among the Andronovans. It is attested by clay 
wheel models, chariot remains, bone cheek-pieces, both complete skeletons and 
the bones of draught animals, and also the images of vehicles and chariots on the 
petroglyphs of Kazakhstan and Central Asia as well as on an Andronovo vessel. 
Two types of vehicle are known: heavy carts with solid disc wheels and light 
single-axle chariots with spoked wheels. 

Vehicles 

At the settlements of Ushkatta 2 (Kuz'mina 1962b, fig. 3), Kambulat (Chebako- 
va 1975, fig. 5.5), Shortandy-Bulak (Margulan 1979, fig. 159.3, 232.11) archae- 
ologists uncovered solid clay wheels with protruding hubs. Such wheels were 
used for vehicle models found in the Eurasian steppes both in burials earlier than 
Andronovo, such as Catacomb-graves of the second half of the 3rd millennium 
BC (Sinitsyn 1948, fig. 13; Sinitsyn and Erdiniev 1971: 8, tables 5a, 8; Kuz'mina 
1974a: 69; Hausler 1981, fig. 11; Romanovskaya 1982, fig. 4.4), and in later 
Scythian-Sarmatian graves, at Mingechaur (Aslanov et al. 1959, fig. 98-101) and 
at Kerch' (Artamonov 1966, fig. 1; Kozhin 1969b: 92-5; Nechaeva 1975: 11-13) 
which demonstrates continuity of tradition and presupposes such kibitkas among 
the Andronovans as well. This is confirmed by the image of a vehicle with four 
solid wheels and solid body, harnessed to a team of bulls, on a petroglyph from 
Saymaly-Tash. In addition to covered vehicles large open four-wheeled wagons 
were also employed, as can also be seen from the Saymaly-Tash petroglyphs 
(recorded by Yu. N. Golendukhin). 

Both open and covered heavy vehicles with disc wheels originate from 
prototypes that appeared on the Eurasian steppes already in the late 3rd 
millennium BC on sites of the Novosvobodnaya and Novotitarovo cultures of the 
Kuban Pit -grave culture and the subsequent Catacomb culture of the early 2nd 
millennium BC (summary of findings and bibliography in Piggott 1969; 
Kuz'mina 1974a; Hausler 1981; Kozhin 1988; Izbitser 1993). E. V. Izbitser 
recorded about 250 burials with one or two vehicles. 

All remaining early vehicles are four wheeled ones. However, in some graves 
there is a pair of wheels and clay models of single-axle vehicles of Kalmykia 
which presupposes, contrary to the opinion of E. V. Izbitser (1993), the existence 
of two-wheeled vehicles on the steppes. An important innovation of the 
Catacomb period was the appearance of the cross-bar wheel. Such wheels are 
known from the Near East and the image of a two-wheeled vehicle with cross- 
bar wheels is known from the stone covering of a cist from the Novonikolaevka 
cemetery (Rassamakhin 1999: 47, fig. 3, 56). 



108 CHAPTER EIGHT 

Evidence of wheeled vehicles is of great importance for resolving the Indo- 
European problem because there are common names connected with wheeled 
transport in all Indo-European languages. Terms denoting 'circle', 'wheel', 
'vehicle', and 'chariot' go back to the verbs *k w el and *(H)ret- 'revolve'. The 
Indo-Iranian words 'yoke' (yugd-), 'axle' (dksa-), 'shaft' (dhur-), and also 'way', 
'movement' have Indo-European correspondences, from which derive the names 
of the race course and the name of the draught animal specific to each deity 
{ydhana-; Schrader 1913; Meillet 1938; Abaev 1949; Gamkrelidze and Ivanov 
1984; Vasmer 1950-1957; Benveniste 1949; Brandenstein 1962). Transport 
terms are included in the sacred vocabulary: 'solstice', 'year circle', 'wheel-sun' 
(Schmitt 1967: 166-169). There also appears a large layer of common Indo- 
Iranian terms: 'saddle-girth', 'shaft', 'bridle', etc. 

Vehicles evolved from sledges, which were known in Mesopotamia and the 
Tripol'e culture already in the 4th millennium BC. In the steppes the earliest 
examples of vehicles are found in the Kuban on sites of the Novosvobodnaya 
culture, with Near-Eastern connections, and the Novotitarovo culture which was 
in contact with both the Novosvobodnaya and the Pit-grave cultures. 

Early vehicles had two axles with four solid disc wheels, 0.7-0. 85m in dia- 
meter; later the standard diameter became 0.5 -0.6m. The wheels were tripartite 
with a protruding hub, sometimes fastened with laths. They were attached to a 
fixed axle, 2m long; it was rectangular in section in the center, but round at the 
ends on which the wheels were inserted. The shaft was in the front part. Two 
structural forms are known: either a draught pole, 2m long, that branched into a 
V-shape, lm long, or a squared beam 2m long with two perpendicular holes. The 
body consisted of a rectangular frame, 2m long and 1.5m wide, with a wooden 
floor covered with mats; the sides were vertical and often with flat or arched 
overhead cover made of wooden laths or bent and crossed branches as seen in 
models and in some Catacomb vehicles (Kozenkova 1973: 65; Romanovskaya 
1982: 107) and in the carriage from kurgan 5 at Pazyryk (Rudenko 1953, fig. 26, 
27); the entrance was in the front. This type of vehicle is represented by a 
Scythian clay model from Kerch' . 

In detail of construction the steppe vehicles differed from those of the 
Danube (Bona 1960: 83; Bichir 1964: 67), Central Asia (Kuz'mina 1983; 
Littauer and Crouwel 1973b; 1974; 1977) and Harappa (Mackay 1951, table 29). 

Ancient authors such as Hippocrates (On Air, Waters and Regions, 25ff), 
Herodotus (4.121) and Strabo (7.3.17) describe the homes of Iranian-speaking 
steppe nomads as wheeled houses, covered with skins or felt. J. Hertel (1925; 
1931: 277, 279) believed that such large houses (kibitkas) of the steppe nomads 
served as prototypes of the vimana-, the celestial houses of the Indo-Aryan gods 
described in Vedic literature. 

The Andronovans used oxen and heavy draught horses for drawing the 
vehicles. V. I, Tsalkin (1972b: 72) proved that Andronovans and Timber-grave 
people could already geld bulls and employed large oxen 133cm at the withers 
for draught. Andronovans were the first in the Old World who bred special 
heavy draught horses, reaching 166cm and more at the withers; they were the 
tallest horses of the Bronze Age. 

Bactrian camels were also used as draught animals. Their ritual burials are 
known from Aksu-Ayuly 2, Telzhan-Kuzeu, Begazy, their bones are found at 



TRANSPORT 109 

Sintashta, Alekseevka, Atasu, Ust'-Narym, Milykuduk, and a clay figurine 
comes from Ushkatta 2 (Fig. 55: 8; Kuz'mina 1963a; 1962b, fig. 3, 4; Margulan 
et al. 1966: 175; 1979: 258, 259; Krivtsova-Grakova 1948: 102; Chernikov 
1966). Pairs of Bactrian camels harnessed to a vehicle are also seen on petro- 
glyphs (Fig. 34) dated to the Bronze Age at Koybagar and Arpauzen in south Ka- 
zakhstan and Saymaly-Tash in the Tian-Shan (Kadyrbaev and Mar'yashev 1967: 
167ff., fig. 22, 100.3; Agafonova 1977: 13; Medoev 1979, table 16; Kuz'mina 
1980c: 29-30). There are numerous individual camel images on petroglyphs in 
Kazakhstan and Central Asia, but they are impossible to date and there are no 
data on their usage as beasts of burden in the Bronze Age. Outside of the natural 
habitat of the Andronovo culture Bactrian camels are reliably known only from 
the south of Central Asia (Tsalkin 1970b: 136, 157; Ermolova 1976: 111; Kuz'- 
mina 1963a: 38-44; 1980c: 19-27, 34; 1983) and Outer Mongolia. They were 
probably found also in Afghanistan and east Iran (Bulliet 1975: 141, 148; Com- 
pagnoni and Tosi 1978: 87-100). In the Near East the Bactrian camel was known 
only at the very end of the 2nd / beginning of the 1st millennium BC (Walz 
1951; 1954; Bulliet 1975: 153-159). In the documents of Tiglath-pileser I (1116- 
1077 BC) Assur-bel-kal (1074-1057 BC) and Ashurnasirpal I (1050-1032 BC) 
they speak of a two-humped camel and from Salmanasar II, we find that camels 
are brought from the east (Luckenbill 1927; Yankovskaya 1956). In early 
Assyrian texts the Bactrian camel is called a gammalu, a Semitic term for drome- 
dary, but it is said at the same time that it has two humps. In the 1 1th century BC 
the term udru appeared, presumably borrowed from Iranian (Salonen 1955-1956: 
85-87). This indicates that the Bactrian camel was alien to Near Eastern peoples 
and it was adopted by them from Iranians. The name of the camel, *ustra-, is 
common in the Indo-Iranian languages (Schrader 1901: 405; Redard 1964: 155- 
162; Mayrhofer 1956: 113-114; Burrow 1976: 143) and it differs from other 
Indo-European languages where the term to denote camel is a late borrowing 
from Semitic gammalu. It follows from these linguistic data that the Indo- 
Iranians became acquainted with the Bactrian camel when they had already split 
from the other Indo-Europeans but were not yet separated into Indo-Aryan and 
Iranian branches. Palaeozoological data show that this could have happened in 
either Andronovo territory or south Central Asia (Anau culture). The name of the 
Bactrian camel was borrowed into the Finno-Ugrian and Turkic languages from 
Indo-Iranian (Bogolyubsky 1929: 14, 15). In Chinese mythology the Indo- 
Iranian cult of the Bactrian camel was accepted in the Han period; its motherland 
was set to the north-west within the borders of north Central Asia (Schafer 1950: 
269, 275). Taking into account Chinese evidence and the localization of the 
Finno-Ugrians in the forest zone of Eurasia, the Indo-Iranian homeland should 
be placed in the Urals and Kazakhstan, and the hypothesis of a migration from 
the Near East should be rejected because the Bactrian camel was unknown there. 



Chariots of the Eurasian steppe 

Single-axle chariots with spoked wheels harnessed to a pair of horses with the 
help of cheek-pieces were of tremendous importance to the culture of the 
Andronovo people (Fig. 35). 



110 CHAPTER EIGHT 

Chariots are recognized from the imprints of wheels with spokes in the Urals, 
in the Sintashta cemetery, burials 5, 12, 16, 19, 28, 30 (Gening et al. 1992: 130, 
132, 149, 153, 163, 165, 167, 183-4, 203, 205, 209-10, 214-5: fig. 56, 72, 78, 80, 
91,93,94, 106-108, 111, 1 1 6, and maybe in burial 14 of kurgan C-l p. 276); Ka- 
menny Ambar, in graves 6, 8 (Kostyukov et al., 1995: 162-3, fig. 9); Solntse-2 
kurgans 4/1, 5/2, 11/2 (Epimakhov 1996: 26, 29, 33, fig. 4); Krivoe Ozero (An- 
thony 1995; Anthony and Vinogradov 1995), Vetlyanka Kurgan 14/6 (Gorbu- 
nov et al. 1990: 32), in north Kazakhstan, in the burials of Ulyubay, Kenes Kur- 
gan 3/5, Berlik-2 Kurgan 2/1, 10 (Zdanovich 1998: 71-6, 138-140, fig. 29: 2-6, 
31: 10-12), and in central Kazakhstan, in the Satan cemetery, kurgan 1 (Tkachev 
1987: 26; 1991: 10, 11; Evdokimov 1981: 434; Novozhenov 1989: 110-122). 

The diameter of the wheel with a protruding hub and ten to twelve spokes is 
recorded at 0.9-lm at Sintashta to 1-1. 2m at Satan; the distance between the 
wheels (ruts) is 1.2-1. 45m, the rim is 4cm wide, the spokes are rectangular and 
square in cross-section, from 3 x 3.5cm to 3.5 x 4.5cm. The frame measures 1.2 
x 0.67m (or 1.6 x 0.6m); fragments of a rim with round spokes, protruding hub 
and red leather tyre fixed by bone nails has been reconstructed for Satan. 

We do not have enough evidence to reconstruct the chariot type. Suggested 
reconstructions (Gening et al. 1992; Anthony and Vinogradov 1995) have been 
justly criticized (Littauer and Crouwel 1996: 934-939). The difficulty of reconst- 
ruction is made worse by the fact that the chamber size was not sufficient to 
insert a whole chariot into it; it had to be put there in parts, or only certain 
elements were deposited. Thus at Kamenny Ambar, in kurgan 2/8 only one 
wheel was found and one cheek-piece out of four was an imitation (Kostyukov et 
al. 1995: 163, 175, fig. 24; 1). 

Horses, usually in pairs, heads facing one another or laid one after another 
may be found in the grave, but more often on the clay ground above the roof of 
the grave. In the Bestamak cemetery three horses were buried standing up. In 
some cases, according to the principle pars pro toto, only skins with the heads 
and legs or only horse skulls are found. Ritual burials of horses have been dis- 
covered in the Petrovka-type cemeteries of Troitsk, Raskatikha, Aksayman, Ber- 
lik Ulyubay and Nurtay (AO 1979: 433; Zdanovich 1988: 73, 74, 78, 80, 135, 
fig. 29). Horse head-and-hooves burials are found at Alakul' and Kozhumberdy 
type sites in graves from Stepnoe I, II, Spasskoe I, Pyatimary, Orsk, Novy 
Akkerman, Alakul', Kupukhta, Emba, Chaglinka, Sokolovka, Semipalatnoe, 
Bylkyldak, etc. (Kuz'mina 1977: 30). Similar horse burials, often of the head- 
and-hooves type, are found in the Early Timber-grave cemeteries of the Pokrovs- 
kiy and Maevka types on the Volga and in the Ukraine (Kuz'mina 1977; Mikhay- 
lova and Kuz'mina O. 1999). In Sintashta, Kamenny Ambar (Fig. 112: 3) and 
Ulyubay pairs of horses were placed together with chariots; in Sintashta and in a 
number of other cemeteries of the Novokumak type cheek-pieces are preserved. 
However, the figure displayed in Gening (1977: 66, fig. 6.1; Gening et al. 1992, 
pi. 22) is an artistic reconstruction, and the cheek-pieces were not found in situ. 
Cheek-pieces along with pairs of horses were also found at Alakul', Nurtai and 
in the Timber-grave burial at Komarovo. It follows that the horses sacrificed 
were intended to be harnessed to the chariots. Palaeozoologists have concluded 
that they were slender-legged young stallions (Sal'nikov 1951b; 1967; Tsalkin 
1972b: 74-77). 



TRANSPORT 1 1 1 

A complete charioteer's complex is also known: the burial of a chariot, two 
horses and cheek-pieces, are known from the Urals, in Sintashta and Petrovka 
burials. On the Volga only a single wheel from kurgan 6/4 at Utevka is known at 
present. There are burials of pairs of horses at Potapovka in the same region, but 
more often only horse head-and-hooves as well as cheek-pieces are found (Va- 
sil'ev et al. 1995). On the Don I know of only one discovery of a wheel from the 
Pichaevskiy kurgan (Moiseev et al. 1995: 75); there are no paired horse burials, 
they are represented only by the head-and-hooves of animals and cheek-pieces 
according to the principle of pars pro toto. This supports the conclusion that the 
Urals was the center of the invention of the chariot and the associated cult. 

The importance of the chariot for the Andronovans is indicated by numerous 
images on petroglyphs at Sary-Arka, Moinkumy, in Dzhambul city, Balkhash, 
etc. (Medoev 1979, fig. 31; Novozhenov 1989: 101, 102; 1987: 86), Tamgaly- 
Karatau in south Kazakhstan (Maksimova 1958: 108-110, fig. 30, 38; Kadyrbaev 
and Mar'yashev 1977: 162-170, fig. 100, 102, 104), Smagul, Moinak and Talapty 
1 in east Kazakhstan (Samashev 1977: 520; Mar'yashev and Rogozhinsky 1991, 
fig. 1-5), Saymaly-Tash in the Tian-Shan (Bernshtam 1962; Agafonova 1977: 5- 
6; Sher 1978: 163ff), Kopal (Makhmutov 1971: 66), Tekke-Tash and Akdzhilga 
in the Pamirs (Mandef shtam 1961: 86, fig. 1, 4; Zhukov and Ranov 1972: 540, 
fig.; 1974: 62-8; Fig. 34). Chariots were harnessed to a pair of horses, with 
wheels of 4, 6 or 8 spokes with rectangular, oval or semi-oval body situated on 
the crossbeam and axle but more often shifted before the axle. Chariots are 
depicted as if in plan, with wheels sprawled along the sides of the body; the 
horses are shown one above the other or against each other, often back to back as 
they are in burials. Possibly this style of depicting wheeled transport appeared 
due to the custom established in the steppes from the Eneolithic to remove the 
wheels from the axle and place them in the corners of the grave (Bussagli 1955). 
The depiction of wheeled transport in plan is characteristic of the large north 
Eurasian zone stretching from Scandinavia to Mongolia (Bussagli 1955; Anati 
1960; Kozhin 1966: 61). In another zone of the Old World, in the Near East, 
chariots were usually depicted in profile. They are represented in this manner on 
seals and reliefs (Nagel 1966: fig. 15-30). Chariots are depicted in the same way 
on Mycenaean stelae, paintings, seals, and ceramics (Karo 1930: tables 5, 7, 10, 
24; Anati 1960, fig. 3, 4; Furumark 1941, fig. 56) and in Mycenaean tablets 
(Wiesner 1998, fig. 5, 6, 13; Lurie 1957: 248, 334, fig. 45). A chronological 
explanation for these differences (Sher 1980: 202-5) does not seem convincing. 
Probably, these are two different artistic traditions (Kozhin 1966: 81; Kuz'-mina 
1980c: 32-34). With respect to stylistic features Kazakhstan and Central Asian 
chariots are especially close to the images on the petroglyphs of the Altai, 
Xinjiang, Tuva and Mongolia, constituting a single Central Asian province 
(Dorzh and Novgorodova 1976; Novgorodova 1978: 203-206; Devlet 1976b: 28; 
Sher 1980: 232). 

Great success has been achieved in resolving the chronological problems of 
the petroglyphs. A considerable number of the chariots can be dated to the 
Bronze Age on the basis of this evidence: 

1) Chariot images analogous to the ones on petroglyphs (Fig. 34: 6, 8) have 
been discovered on vessels from the Alakul' burial at Spasskoe and the Timber- 
grave Sukhaya Saratovka 2 (Stokolos 1972: fig. 13,2, Galkin 1977: fig. 1), and 



112 CHAPTER EIGHT 

on Timber -grave pots from Politotdel'skoe, from the Lvov and Zhdanov museum 
wheels with axles are depicted (Kuz'mina 1974a: 82; Cherednichenko 1976: 
139, fig. 4-6). These can be correlated with European data. Depictions of 
chariots from Kivik are dated to the Bronze Age according to its tumulus 
complex (Glob 1974), a number of images are found on vessels of the Bronze 
Age (Vizdal 1972: 233, fig. 1; Cherednichenko 1976: 139, fig. 4-6; Smirnov and 
Kuz'mina 1977: 54-55; Piggott 1978: fig. 1). In the Caucasus there is representa- 
tion of vehicles in plan, including a Bronze Age vessel from Dilizhan in Armenia 
(Esayan 1976: table 97, 1) and also a bronze model of a two-wheeled chariot 
with ten-spokes from Gokhebi in Georgia (AO 1976: 478, 479). Caucasian exam- 
ples are important because some names relating to wheeled transport are com- 
mon to the Indo-European and Caucasian languages, including Nakh-Dagestani. 

2) Palimpsests are known on petroglyphs. Chariots are covered over by 
Scythian drawings (Mar'yashev and Rogozhinsky 1991: 6). 

3) In Tuva chariot images are found in combination with Okunevo 
escutcheons (Devlet 1976 b: 28). 

4) Two stylistic groups of chariot images are found in Mongolia, the later is 
dated on the basis of a Scythian weapon on a stone from Darvi-Somon, which is 
set to the Scythian period and serves as a terminus post quern for the earlier one 
(Novgorodova 1978: 203-206). 

5) The symbol for the chariot in the early Chinese script is stylistically close 
to chariot images on Central Asia petroglyphs (Fig. 54: 13). This analogy is 
chronologically important because some researchers of wheeled transport believe 
that the chariot was borrowed into China from the Eurasian steppes. Chinese 
examples and chariot images on petroglyphs help partially to reconstruct the 
more ancient type of Andronovo chariot. It likely had a single axle carriage with 
draught-pole and two forked yokes to which horses were harnessed. Rotating 
wheels were attached to a fixed axle; then onto the axle and pole the frame was 
set; the frame was sometimes shifted forward in front of the axle. The chariot 
size was smaller than that of the Shang. 

As for the wheeled transport of the ancient Aryans in the Near East in 
Mitanni, it is reconstructed on the basis of written sources (Zaccagnini 1977: 21- 
38). In texts from Nuzi written in the Hurrian language but preserving Indo- 
Iranian names connected with transport and the horse, heavy transport vehicles 
are contrasted with lighter battle and ceremonial chariots. The Mitanni king 
Tusratta sent a whip with a gold knobbed handle decorated with lapis-lazuli to 
the Egyptian Pharaoh Amenhotep III. 

The wheeled transport of the Aryans of India is reconstructed on the basis of 
Vedic texts and the Mahabharata (Kal'yanov 1967: 136-160; Kane 1946: 200- 
205; Gode 1960; Pusalker 1963; Margabandhu 1973; Sparreboom 1985; see Fig. 
95: 2; Rau 1986; see Fig. 95: 2). The Vedic Aryans used a draught vehicle anas- 
and a battle chariot (rdtha-). Draught vehicles were of two types: a large firm 
four-wheeled (indranasa-) and the more ordinary two-wheeled vehicle with a 
fixed axle (dksa-), attached to the body with ropes. The wheel was composed of 
three parts: two outer segments and a central part with a protruding hub, which 
were fastened by cross laths {Satapatha-Brahmana 5.4.3; 7.2.3). "Three planks 
form the vehicle wheel. There is a large hole for the hub of the central plank to 
insert the axle into it. . . A pair of side planks is fixed by bushing keys" (Aitareya- 



TRANSPORT 113 

Brdhmana 4.15.6). The wooden draught pole was perpendicular to the axle and 
above it; a double pole was fixed with wooden nails. From behind and under the 
body there was a loop used probably as a brake. Following the text of the Kdtyd- 
yana-Srautasutra (8.4.5), "the wheels of both draught vehicles are tri-partite. 
They are fixed on both sides of the axle... There are also two poles (i.e., a double 
pole) extended forward; an axle; cross-beam in front and behind separating both 
parts of the pole; ... the yoke is attached to a forked pole... This was the const- 
ruction of both vehicles". The yoke (yugdm) was attached to the front part of the 
draught pole by two depressions for the back of the head of the draught animal. 
The yoke was fastened to the pole with ropes; there were four holes in the yoke 
into which wooden pegs were inserted. The pegs were on both sides of the 
animal to fix its head. From the yoke the two ends of the double pole met to form 
an isoscelese triangle. In the Rigveda (3.53.17) Indra is called: "You whose 
wheel-rim endures no damage, be of help to us! Both oxen must be hardy, and 
the axle must be firm. The pole must not unfasten or the yoke must not break. 
Indra must preserve both vehicle supports from breaking." The frame of the 
vehicle was planked and above it were bent arched planks; it was re-covered by a 
'cap' (gadhd-) made of mats or pieces of felt. A door was made in the edge of 
the kibitka. The body of the vehicle was 'a hut for women', 'a hearth', the gdrha- 

patya the interior of the vehicle; there is a "forked sharpened pole, a yoke, a 

step into the body" (Rau 1986: 18, 24). 

In later texts six types of vehicles are mentioned: those for carrying people 
(drohana-), for the hearth and kitchen utensils {mahdnasa-) and special vehicle 
types for carrying military equipment and chariots. Using three-spun ropes first 
the left then the right-hand animal was harnessed; halters with reins (rasmi-) 
were attached. The priest exclaimed: "You belong to Varuna! The rope of 
Varuna is set" {TaittirTya-Samhitd 1.2.8.2). 

Oxen, bulls, horses, mules, and the ceremonial harnessing of a horse and a 
mule were used. Vehicles harnessed to four camels are mentioned in the Rigveda 
(8.6.48) and later texts. 

Chariots were two-wheeled. Detachable wheels had an indefinite number of 
spokes, rim and hub; sometimes they were bound with metal. The axle was 
attached to the frame by ropes, the pole was straight. There is a vague mention- 
ing of a chariot one horse high. In the Rigveda (4.40) a trotter is described with a 
bridle on its head and harness on the muzzle. The harness ran from behind the 
back of the head to attach to the saddle-girth. In the Satapatha-Brdhmana and 
other texts the halter is mentioned many times. The driver bore a whip. In the 
Jaiminiya-Brdhmana (1.129-130) two horses, poles, reins, both wheels, and the 
body of a divine chariot are mentioned. In the Taittinya-Brdhmana (1.5.12.1) the 
two leading and two interchangeable trace-horses of Prajapati's chariot are 
described; the chariot "gleams and it is decorated with metal plates and balls", it 
has a skin-made cover, sections for quiver and arrows (gorytus). The driver and 
warrior in a leather coat of mail are in the chariot. "Indra is the chariot warrior, 
the moon is the driver" (Atharvaveda 8.8.22-23). The chariot descriptions with 
warrior and driver are numerous in the Mahdbhdrata. 

According to Vedic texts they made semi-subterranean sheds for keeping the 
chariot with ditches for placement of the wheels (Rau 1986; Renou 1939: 491, 
492) as in the Sintashta graves and in the Chinese che-ma-ken. 



114 CHAPTER EIGHT 

According to W. Rau the length of the chariot axle of the Vedic period was 
2m, the pole length 3.61m, the yoke 1.65m. As far as descriptions are concerned, 
the Aryans applied a dorsal-yoke harness system, reproduced in modern experi- 
ments, in which a saddle-girth belt is used connected to the horse's collar; the 
main effort is not on the animal's neck (as with the throat harness for oxen) but 
on its chest (Spruytte 1977: 13, 17, fig. 14, 20, 21, table 16). This type of harness 
is preserved in modern India. There is not enough data to identify the harness 
type of the Andronovans. But chariot construction according to the images on 
petroglyphs can be compared with the Vedic evidence. A graphic reconstruction 
of the Vedic chariot on the basis of textual evidence was suggested by S. Piggott 
(1962: 32; Rau 1983, fig. 5). Petroglyphs of north-west India in Gilgit (Thor) 
(Fig. 34: 15), populated by relict Indo-Iranian speakers who left their homeland 
before the Indo-Aryans, represent the two-wheeled horse-drawn chariot (Jettmar 
1985: 755, 757, fig. 6). In central India there is an image of a two-wheeled two- 
horse chariot (Lai 1961: fig. 5, 6); in Mirzapur we find four-wheeled vehicles 
harnessed to a pair of horses (Kasambi 1968: 123); in Merkhan Pakhar (Mirza- 
pur) there is a two-wheeled chariot with four horses (Allchin and Allchin 1973, 
fig. 4); on a vessel of the Jorwe type from Inamgaon there is a two-wheeled ox- 
cart drawn by a pair of bulls (Shchetenko 1979, fig. 39). These images are not 
executed in the Near Eastern manner, in profile, but they are made in plan, in the 
style characteristic of northern Eurasia, Kazakhstan and Central Asia. This 
important stylistic feature reflects the Andronovo influence on the development 
of wheeled transport in India in the second half of the 2nd millennium BC. 

It is probable that these steppe tribes were decisive in the development of 
horse-breeding in India as well. Although the domestic horse was already known 
to the Harappan culture, only in the post Harappan period in India was horse 
breeding and the horse cult documented by ritual horse burials or their images in 
Swat (Katelai) and Gomal (Gumla, Hathala; Dani 1970-1: 49-53; Castaldi 1968: 
598-602). This rite has no sources in India, south Central Asia and Iran, or the 
Near East but it is well known on the Eurasian steppes from the Eneolithic and 
was especially popular among the Andronovans (Kuz'mina 1974a; 1977: 34). 
The possible influence of steppe tribes, coming from Central Asia, on the devel- 
opment of horse breeding in India is affirmed by paleozoologists who have de- 
duced that the Swat horses belonged to the eastern breed (Azzaroli 1975: 353-5). 

The appearance of the Bactrian camel as a draught animal in Baluchistan is 
probably connected with Andronovo influence. On the post-Harappan settlement 
of Pirak camel bones and clay camel figurines (Fig. 104), including figurines 
with holes for attachment to vehicles, and horses and mounted horsemen were 
found. J. F. Jarrige (1979: 92-5, 100) dates the complex to the second half of the 
2nd / beginning of the 1st millennium BC. He remarks that Bactrian camels are 
not characteristic of the previous Baluchistan cultures and their appearance in 
Baluchistan is connected with Central Asian influence. The Bactrian camel was 
not known in Harappa but it often appears in Vedic literature. The role of the 
Bactrian camel in the economy led to the formation of its cult, brilliantly ref- 
lected in Indo-Iranian religious tradition. In the Rigveda (8.6.46) there is mention 
of four pairs of camels (yoked), and camels are mentioned in the Mahabharata 
and in the Laws ofManu (3.162; 4.115,120; 5.3; 8.296). The camel was especial- 
ly honored in the Avesta, where it appears many times and comes as the main 



TRANSPORT 115 

sacrifice animal (Videvdat 9.37, Yasna 44), as one of incarnations of the wind 
god Vayu, the storm and victory god Vsrsthragna (Bahrain Yast, 8), and the 
embodiment of Glory-Farnah (Zamyad Yast, 34-38). 

Cheek-pieces 

The Andronovans employed cheek-pieces of horn and bone to harness horses to 
their vehicles. These can be divided into three classes, shieldlike, grooved and 
rodlike. The shieldlike cheek-pieces are subdivided into two groups, those with 
tenons and those without. 

Shieldlike cheek-pieces with tenons 

A large number of these cheek-pieces are found on sites of the Sintashta-Petrov- 
ka type (Figs 36; 37; 39), in the Urals — 15 various examples from the cemetery 
of Sintashta I (Gening 1975a: 94), one cheek-piece from Sintashta II (AO 1975: 
168), one from the settlement of Kulevchi III (AO 1979: fig. on 138), in northern 
Kazakhstan on the settlements of Novonikol'skoe (fragment) and Petrovka II 
(one and three half- finished) and four from the cemetery of Berlik II, kurgan 10, 
together with a chariot (Zdanovich and Khabdulina 1976: 96; Zdanovich 1983: 
57, 63, fig 3.1, 2, 7, 9, 10; 1985; 1986; 1988: 138-139, fig. 31.7, 8). 

The Petrovka cheek-pieces belong to a category that was distributed widely 
over the Eurasian steppe (Map 7). It was A. M. Leskov (1964) who examined 
them and first mentioned their similarity to objects from Mycenae followed by 
B. A. Latynin (1965), A. D. Pryakhin (1976: 122-124, fig. 2), K. F. Smirnov and 
E. E. Kuz'mina (1977: 42-43, fig. 11). Analogous cheek-pieces are known from 
the Danube region (Oancea 1976: 59-75; Huttel 1978: 65-86). 

In 1978 the author developed a scheme for the evolutionary development of 
the types of cheek-pieces of the Eurasian steppe and neighboring regions on the 
basis of the typological-technological method (Fig. 37; Kuz'mina 1980: 8-21, 
fig. 1) which has received some recognition (Azzaroli 1985: fig. 10). It has stood 
the test of time and may be a significant addition to the results of the most recent 
discoveries of cheek-pieces in the cemeteries of the Abashevo culture, 
Tavlykaevo IV in the Urals (Gorbunov 1986: tab. xvii, 6); Utevka IV on the 
Volga, Vvedenka on the Don (Vinnikov and Sinyuk 1990: 127) and especially in 
the burials of the chariot -warriors in the kurgans of the Potapovka type, 
contemporary with the closely related Petrovka, Potapovka, Utevka VI on the 
Volga (Vasil'ev et al. 1992; Kuznetsov 1993: 74), Kondrashkinskiy I (Pryakhin 
1992) and Pichaevskiy (Moiseev 1990: 56) on the Don (Table 5). 

Type 

IA 



Culture 


Site 




Tenons 


Panel 






number 


type 


shape 


Late Helladic 


Mycenae, grave 4 


3 




solid 


round 


Petrovka 


Petrovka II 


4 




solid 


round 


Sintashta 


Sintashta, graves 5 


, 










11,30,39 


4 




solid 


round 


Timber-grave- 












Abashevo 


Surush 


3 




solid 


rectangular 


Abashevo 


Tavlykaevo IV 


3? 




solid 


round 





CHAPTER EIGHT 






Abashevo 


Balanbash 3 


solid 


round 


Catacomb- 








Multi-roller 


Kamenka 2 


solid 


round 


Late Helladic 


Mycenae, grave 4 3 


solid 


round 


Sintashta 


Sintashta, grave 12 4 


solid 


round 


Potapovka 


Potapovka 3/4 4 


solid 


round 


Multi-roller? 


Trakhtemirovo 4 


solid 


round 


Monteoru IIA-IIB 


Candesti 3 


solid 


round 


Monteoru IC 4 -IIA 


Carlomanesti 3 


solid 


round 


Monteoru IC4-IIA 


Sarata-Monteoru 3 


solid 


round 


Sintashta 


Sintashta, graves 11, 








12 4 


solid 


round 


Potapovka 


Potapovka 5/8 3 


inserted 


round 


Potapovka 


UtevkaVI, 6/6 3 


1 insert; 

2 solid 


round 


Potapovka 


Kondrashkinskiy 13? 


solid 


round 


Potapovka 


Pichaevskiy 3 


insert 


round 


Timber-grave- 








Abashevo 


Otrozhka 1 


solid 


rectangular 


Abashevo 


Kondrashevka 3 


insert 


round 


Petrovka 


Berlik 10 2 


solid 


oval 


Sintashta 


Sintashta, grave 30 2 


insert, 
solid 


oval 



Potapovka 


UtevkaVI, 










kurgan 2 


2 


insert 


oval 


Potapovka 


UtevkaVI, 










Kurgan 2 


4 


solid 


oval 


Potapovka 


UtevkaVI, 6/4 


4 


solid 


oval 


Potapovka 


UtevkaVI, 6/6 


2,4 


solid 


oval 


Potapovka 


Bogoyavlenskiy 


4 


insert 


rectangular 


Late Abashevo 


Staroyur' evo 


4 


insert 


round 


Coslogeni? 


Ulmeni 


4 


solid 


round 


7 


Sarata-Monteoru 


4 


solid 


round 


Late Helladic 


Dendra, grave 7 




solid 


round 


Late Helladic 


Mycenae, grave 814 


insert 


round 










bronze 


Late Helladic 


Mycenae, grave 15 4? 


insert 


round 


Late Helladic 


Mycenae, palace 




insert 


round 


Late Helladic 


Kakovatos, 










tholos A 


4 


insert 


round 




Tiszafured 


4 


insert 


round 


Abashevo 


Vesely 


4 


insert 


round 



Table 5: Classification of shieldlike cheek-pieces with tenons 



TRANSPORT 117 

Eurasian bone and antler cheek-pieces constitute a single category of artifact 
possessing a flat tenon with a large central round, rarely rectangular, opening, 
sometimes surrounded by a bevel, and three or four (rarely two) protruding 
tenons on the reverse side. According to the form of tenon they are divided into 
three variants: disc, oval and segmented. According to the arrangement of the 
additional openings they are divided into six types. 

Type I, subtype A. Disc cheek-pieces with one or several small openings on 
the periphery of the disc with three or four tenons. These are found on the settle- 
ment of Petrovka II (1 whole and 3 unfinished, made from deer antler), at Sinta- 
shta in burials 5, 11,30 and 39, and also in the Urals on the Abashevo settlement 
of Balanbash (Sal'nikov 1967: fig. 8, 19) and in the cemetery of Tavlykaevo IV 
(Gorbunov 1986: tab. xvii. 6), on the Timber-grave -Abashevo settlement of 
Surush on the Middle Volga (Vasil'ev 1975: fig. 6.2); in the late Catacomb (or 
Multi-roller Ware) settlement of Kamenka in the Crimea (Rybalova 1974: fig. 
3.1; Leskov 1964: 12). Cheek -pieces of subtype Al with three tenons are found 
in grave 4 of Grave Circle A (according to G Karo, the 4th shaft grave) at 
Mycenae in Greece (Karo 1930: 130, tab. 70, no. 532-535). The Mycenaean 
cheek-pieces were originally interpreted as pot covers (H. Schlieman), helmet 
attachments for plumes (Karo 1930: 372), as model thrones (Wace 1960), and as 
votive tablets (Mylonas 1957: 103); their utilization as cheek-pieces was rejected 
because among the pairs of tenons there were transparent holes in which tubes 
were inserted (Littauer and Crouwell 1973a: 207-213). But this does not exclude 
their use as cheek-pieces (Oancea 1976: 71; Htittel 1978). The utilization of disc- 
shaped cheek-pieces was supported by figures from the Tiryns frescoes (Fig. 66: 
8, 10; Evans 1927: figs. 810, 811; Wiesner 1968: fig 15a and b). 

To subtype IB belong two cheek-pieces with three tenons from shaft-grave 4 
with small extruding semi -oval slats without functional significance and 
ornamented on the end of the disc with running wavy lines or running spirals 
(Karo 1930: pi. 70, no. 533, 534). 

Subtype IC is defined by disc-shaped cheek-pieces with large central holes, 
sometimes with one or two small ones, and several small holes on the end of the 
disc. They are from Sintashta (grave 12), Potapovka (kurgan 3, grave 4) and Tra- 
khtemirovo in the Ukraine. A. M Leskov (1964: fig. 1) and S. S. Berezanskaya 
(1982: 31, fig. 40.1) suggest a connection between the latter and the Multi-roller 
Ware culture. 

The construction of the type II cheek-pieces is distinguished by distinctive 
slats on which small holes are distributed. To subtype IIA, those with a rectangu- 
lar slat, belong those cheek-pieces with a truncated segmented disc and arranged 
along the chords are small holes, often with inserted tenons. These are found at 
Sintashta, graves 11 and 12, at Potapovka, kurgan 5, grave 8 (Kuznetsov 1993: 
74), Utevka VI, kurgan 6, grave 6 (Vasil'ev et al. 1992: fig 5, 12), in the kurgans 
at Kondrashkinskiy I (Pryakhin 1992: fig 1.1), Pichaevskiy (Moiseev 1990: 56), 
in the Abashevo cemetery of Kondrashevka (Efimenko and Tret'yakov 1961: fig. 
31.4), and on the Timber-grave -Abashevo settlement of Otrozhka (with thanks to 
B. G. Tikhonov for the unpublished information). Three cheek-pieces of subtype 
IIA have been found in the Danube region, a fragment of one in the ashpit about 
the settlement of Carlomanesti along with ceramics belonging to Monteoru IC 4 - 
IIA; a second from the settlement of Sarata-Monteoru in a complex of the same 



118 CHAPTER EIGHT 

culture (Oancea 1976: fig. 3, 162); a third from the settlement of Candesti in a 
level belonging to Monteoru IIA-IIB (information from V. S. Bochkarev). 

To subtype IIB belong cheek-pieces with tenons of an oval (rarely rectangu- 
lar) form, with a central hole and with a markedly protruding triangular tag with 
additional holes. Such cheek-pieces are known from Sintashta, grave 30, Berlik 
II, kurgan 10 (Zdanovich 1988: fig. 31.7,8), and also along the Volga Utevka VI, 
kurgan 2 and 6, graves 4 and 6 (Vasil'ev 1975: fig. 6.1; Vasil'ev et al. 1992: fig. 
4.7-9, 18, 5.13, 6.9-11) and on the Don from the cemetery of Bogoyavlenskiy, 
kurgan 1, grave 3 (with thanks to Yu. P. Matveev for this information). 

A special variant of the cheek-pieces consists of examples of segmented 
shields, with a large central hole and tenons. They are found on the settlement of 
Novonikol' skoe (Zdanovich 1983: 3.1; 1988: pi. 10.9), Kamyshnoe (Potemkina 
1985: fig. 107, 192; dates of the 15-14th centuries BC are not settled) and 
Kulevich III (AO 1979: fig. on p. 138). Possibly we can add here a bone object, 
found on the Timber-grave settlement of Sosnovskiy 3 in the Samara region, as a 
similar unfinished cheek-piece (information from S. A. Agapov). Unfortunately, 
all four examples are fragmentary so that they do not permit complete identific- 
ation. It is possible that they had the distinctive tag with the small hole and may 
be assigned to subtype IIC. 

To type III belong cheek-pieces having a distinctive rectangular tag with 
small holes and additional holes on a second flat piece which principally distin- 
guishes them from subtype IIA. There are two cheek-pieces from the cemetery of 
Staroyur'evo on the Don having four insertions, pegs, ornamented on the outside 
with crosses, running zig-zags and with teeth along the edge of the disc 
(Pryakhin 1972: fig. 3). 

To type IV belong the cheek-pieces with a large central and two smaller 
holes on the periphery of the disc and additional transverse holes with four solid 
tenons. So far they are known only from the Danubian region, one from Ulmeni 
which was found together with ceramics of the Cernavoda (Eneolithic) and 
Coslogeni (Bronze Age) cultures; a second is from Sarata-Monteoru from an un- 
known context (Oancea 1976: fig. 5.34). The first cheek-piece was ornamented 
along its ends with zig-zags; the second with two rows of teeth. 

Type V cheek-pieces are round discs; in contrast to the preceding they have 
two equal central holes (sometimes with bevels) and on the four inserted pegs 
(sometimes decorated with grooves), inserted in the rectangular holes of the disc. 
They are so far known only from Greece, three or four cheek-pieces were found 
in the palace of Mycenae in the 'House of Shields'; fragments of two more were 
found at Mycenae, possibly in grave 15; one come from tholos A at Kakovatos 
on the Peloponnese, one is from Dendra, grave 7 (Wace 1960: figs. 2-4; Oancea 
1976: figs 4.5-10). To the same type one should add the bronze disc with two 
central holes and four holes for receiving the dowels. It was found at Mycenae in 
grave 81 (Fig. 67:4). 

Type VI consists of disc-like cheek-pieces with protruding plates on which 
additional holes are situated. Two examples with four tenons and a small slat 
with two small holes were found by M. S. Smirnov in 1920 in kurgan 1 at the 
site of Vesely near Sterlitamak in Bashkiria in a complex with pottery with 
crushed shell temper (Archive of the St Petersburg Institute of Material Culture 
63, 1920: 2-4, with thanks to V. P. Shilov for making me aware of this find). 



TRANSPORT 119 

Analogous cheek-pieces with several holes in the disc are from Tiszafured in the 
Danubian region (Bona 1975: tab. 195.1). 

In addition to this, in central Kazakhstan A. A. Tkachev (1991: 8, 12, 15) 
refers to finds of shieldlike cheek-pieces along with the bones of a horse; in enc- 
losure 7 of the cemetery of Atasu (Ayshrak) an inserted bone tenon was found in 
an assemblagte with early Alakul' ceramics and a bone bead (Archives of IIAE 
AN KazSSR, Diary of Central Kazakhstan Expedition, 1952). These finds 
indicate the eastern limits of the distribution of disc-shaped cheek-pieces with 
tenons in the Eurasian steppe. 

Analysis of the details of the classification of types of disc-shaped cheek- 
pieces permits us to reconstruct the evolution of the objects of this category (Fig. 
37). Original were cheek-pieces of subtype IA which were distributed across the 
Eurasian steppe from the North Pontic to northern Kazakhstan and at Mycenae in 
Shaft-grave 4. The similarity of the objects of this subtype over the entire area 
and especially the similarity of details not functionally determined leaves no 
doubt that there was a single origin for the disc-shaped cheek-pieces. The most 
archaic in the series are the cheek-pieces of Balanbash which are crude, 
asymmetric, with three tenons and a single small hole. 

An important innovation in the form of disc-shaped cheek-pieces was the 
making of additional holes in the second plate. Along the Danube it was realized 
in Type IV (Ulmeni) which was developed from subtype IA. In south Russia the 
principal new type of cheek-pieces was Type III (Staroyur'evo), which also 
developed from Subtype IIA. In Greece evolution took a different course; on the 
basis of Subtype IA there was created Type V (Kakovatos) which stimulated 
imitations in metal. It is possible that distant replicas of cheek-pieces of Type VI 
with holes on the plate appear as bronze cheek-pieces with tags with holes, as 
were found in the citadel of Mycenae in the Late Helladic IIIB or IIIC level 
(Crouwel 1981: tab. 1). 

The second important improvement in the form of disc-shaped cheek-pieces 
was the application of the inserted peg, which replaced the quickly worn out 
solid tenon (characteristic of the cheek-pieces of Type I) with a type that 
significantly extended the life of the cheek-piece. In the south Russian steppe 
this innovation was already known from objects of Type II and was character- 
istic for the insuing line of evolutionary development of Type III (Staroyur'evo) 
which was similar to the Late Mycenaean cheek-pieces of Type V (Kakovatos). 

Keeping in mind that the genetic line tends to reflect the general tendency of 
development of the cheek-pieces, this does not exclude the mutual co-existence 
of old and new types in a time of intensive search for innovations. In a closed 
context at Sintashta in grave 30 there were found cheek-pieces of subtypes IA 
and IIB, in grave 12 there were subtypes transitional between IC and IIA, at 
Utevka in kurgan 6, grave 6, were subtypes IIA and IIB. But the regularity of 
evolution supports, in the first place, the correlation of changes in form and 
ornament, and secondly it is supported by stratigraphy, e.g., at Potapovka it has 
been established that kurgan 3 with a cheek-piece of subtype IC according to our 
classification was older than kurgan 5 with a cheek-piece of subtype IIA which 
bore Mycenaean ornament (Kuznetsov 1993). 

What about the absolute chronology of the various types? A. M. Leskov 
(1964) dated the disc-shaped cheek- pieces of the south Russian steppe after 



120 CHAPTER EIGHT 

analogous ones from shaft-grave 4 at Mycenae to the 16th century BC. V. A. 
Safronov (1966: 189-190), on the other hand, argued that "cheek-pieces, similar 
to those from Kondrashevka, were distributed in Greece from the middle of the 
2nd millennium BC through the 13th century BC, while they were imported 
there from Egypt... It follows that their appearance in the Don region does not 
require a date earlier than the 13th century BC." The publication of the cheek- 
pieces from the Danube by A. Oancea cites different dates from a variety of 
archaeologists but he does not establish his own chronology. 

The classification of cheek-pieces and the establishment of their evolution 
permits us to establish the origin of the disc-shaped cheek-pieces and their chro- 
nology. The most archaic disc-shaped cheek-piece was amorphous and undecor- 
ated of Type I and derived from contexts of the Catacomb-Multi-roller Ware and 
Abashevo cultures from the Ukraine to the Urals. This permits us to attribute the 
first controlling of chariots with cheek-pieces to tribes of the Abashevo and 
Multi-roller Ware cultures (KMK=Kul'tura Mnogovalikovoy Keramiki). The 
second played a decisive part in the origin of the closely related sites of the 
Potapovka type from the Don to the Volga, and the Sintashta and Petrovka 
cultures of the Urals and Kazakhstan. Among the Potapovka and Sintashta- 
Petrovka population the tactics of chariot warfare underwent a massive develop- 
ment that is evidenced in the first place by the burials of warriors, preserved with 
chariots and horse gear, and, secondly, by the greatest concentration of disc- 
shaped cheek-pieces in the Old World in the forest-steppe between the Don and 
Urals; thirdly, there is the intensive quest here for the most rational construction 
of the cheek-piece, evidenced by the coexistence of various types. Taking into 
account the monocentric origin of the disc-shaped cheek-pieces it is possible to 
suggest that they expanded into Kazakhstan, then into the Danube region and 
further into Greece from the south Russian steppe. In this case the date of the 
cheek-piece from grave IV at Mycenae, belonging to the end of the Middle 
Helladic/beginning of the Late Helladic period, serves as a terminus ante quem 
for the cheek-pieces of Type I from Potapovka, Sintashta and Petrovka, and 
further the most archaic examples of the Abashevo (Balanbash, Tavlykaevo, 
Surush) and the Catacomb-Multi-roller Ware (Kamenka, Trakhtemirovo), which 
may belong to the 17th century BC. 

The development of cheek-pieces of Type II (with the characteristic plate and 
often with inserted tenons), to which belong the majority of the Potapovka and 
Sintashta examples, is established on the basis of synchronization with the 
Danubian examples dating to the period of Monteoru IC 4 -IIA. Subtype IIA and 
IIB are contemporary as indicated by their mutual existence at Utevka VI. 

Some of the cheek-pieces of this subtype are richly ornamented (Figs. 36, 
37). An example from Potapovka is decorated with Mycenaean ornament of 
opposed triangles, forming zigzags to the end of the disc, waves with a recessed 
background over the plate and with festoons about the hole. On a cheek-piece 
from Utevka on the end of the shield there are three rows of checkerboard 
ornament; running waves with recessed triangular foundation are along the plate. 
On the Kondrashkinskiy cheek-piece, both around the hole and on the plate, 
there are festoons and zigzags on the plate that are similar to those from 
Potapovka. The example from Pichaevo is decorated with a chain of rhomboids 
and zigzags on the end, V-shaped figures are on the plate, around the holes they 



TRANSPORT 121 

form rosettes; a cross is on one of the tenons, on the other a four -pointed solar 
symbol. 

The Bogoyavlenskiy cheek-pieces are ornamented on the end plate with a 
running wave and on the third tenon by spirals forming a rosette. The cheek- 
pieces from Candesti are decorated on the end by two rows of triangles forming 
zigzags. On the cheek-piece from Sarata-Monteoru the plate is decorated with 
zigzags and isoceles triangles with festoons hanging from the top; in the center 
of the example from Carlomanesti there are running waves with dots and teeth 
on the end. 

These ornaments belong to the Mycenaean type (Figs. 38; 66; 67; 79) about 
which we have already written (Smirnov and Kuz'mina 1977: 46-50, fig. 12). 
The Bogoyavlenka and Utevka cheek-pieces are ornamented with classical 
Mycenaean running waves. They appear in the art of the Middle Helladic period 
(Mylonas 1957: figs. 43, 45) and are especially characteristic for the time of the 
shaftgraves, cheek-pieces of Type I and plates of shaft-graves III and IV, stela 
no. 1428 (Karo 1930: tab. v, lxx, nos. 533, 535, xx, no. 38). On the example 
from Bogoyavlenskiy the waves are combined with rosettes, which are similarly 
represented on plates from shaft-graves IV and V (Karo 1930: No. 319, 337, 344 
and 676), and at Utevka with the checkerboard pattern, which are figured in 
composition with waves in paintings in the megaron (Mtiller-Karpe 1980: tab. 
243.6). The cheek-pieces from Pichaevo were covered with decoration, created 
out of V-shaped figures, that was very typical for shaft graves III, IV and V 
(Karo 1930: tab. xxi, N 23; lxi, N 340, 345). Similar to the Bogoyavenka five- 
petalled rosette are the bone plates from Ftizesabony and Nitriansky Hradok, the 
four-petalled ones from Veterovo, and the eight-petalled from Tell-Acana 
(Mtiller-Karpe 1980: tab. 291, B8; 294, D7; 298: 17; 159, A4) (similar but 
simplified ornament is known from a plate from the cemetery of Balakty in 
central Kazakhstan). According to the Mycenaean analogies the cheek-pieces of 
Type II may be dated to the 16th century BC. 

To this same time may also belong the bone pendant from the Pokrovskiy 
cemetery, decorated with simple waves, similar to those on Mycenaean cheek- 
pieces of Type I (Karo 1930: tab. 70, 534). It is interesting to analyze the head- 
dress from the Abashevo kurgan of Algashi (Efimenko and Tret'yakov 1961: fig. 
14). The ornament of this, internally filled zigzags, had a long devel-opment in 
Greece (Furumark 1941: motif 65). The Mycenaean diadems, embroi-dered with 
plates with six-petalled rosettes and double-spiral pendants, recall the Abashevo 
culture's initial headbands (Efimenko and Tret'yakov 1961: 56, figs. 10, 12). 
Similar diadems are found in Egypt at Thebes in graves of the XVIII dynasty and 
in the Near East in Assyria that provides a basic synchronization of cultures from 
the three continents to the 16th century BC (Mtiller-Karpe 1978: 16-18). 

The cheek-pieces from Kondrashkinskiy and Potapovka were decorated with 
opposing triangles, forming zigzags to the end of the disc just like a plate from 
grave V (Karo 1930: no 707, 716, 722) and hanging festoons. The latter is also 
represented on cheek-pieces from Staroyur'evo and on a bone disc from the Aba- 
shevo settlement of Shilov on the Don (Liberov 1980: figs. 1.2, 2.2). The motif 
finds an analogy on bone objects from Vatina; many bone whips and staffs from 
Eastern Europe are decorated with Mycenaean motifs (Fig. 79: 11-35). 



122 CHAPTER EIGHT 

In Transylvania and along the Danube the ornaments, which developed under 
Mycenaean influence in the 16th century BC, continued for a long time and were 
transformed (Bochkarev 1968; Hachman 1957; Tocik 1959; Foltiny 1967; 
Kovacs 1969; Bona 1975; Htittel 1982). The post-Mycenaean period saw a 
characteristic degradation of the scheme with the spread of circular ornament 
with dots. The apogee of this ornamental system reached the Noua culture. On 
the cheek-pieces of Type III from Staroyur'evo festoons framed with dots are a 
late characteristic of Type IV (Monteoru, Ulmeni). 

Thus the absolute date for cheek-pieces of Type I in view of the complete 
similarity of this original form over all areas is set to the 16th century BC 
according to the analogy with the 4th shaft grave at Mycenae, which is assigned 
to 1570-1550 BC (Mylonas 1957: 181; Marianatos 1960: 177), while it serves as 
a terminus ante quem for the most archaic form of the south Russian steppe 
which may belong to the 17th century BC. 

It should be mentioned that in the past years there has been a tendency to 
review the chronology of A. Furumark and advance the antiquity of the Myce- 
naean complex by a century. Dates of the 17th century BC are based, firstly, on 
the time of the eruption of the volcano on Santorini (1645 BC) as established by 
radiocarbon dates; secondly, by synchronisms with the Egyptian and Near 
Eastern complexes in which Mycenaean objects have been found (Dietz 1991; 
Warren and Hankey 1989; Kemp and Merrilleese 1980). The chronological 
relationship of the Mycenean graves has also been revised (Killian-Dirlmeier 
1988) and the absolute development of the Egyptian dynasties. However all 
these hypotheses are questionable. But if the dates of Mycenae were to be 
revised, this would permit one to advance the antiquity also of the Eurasian 
complexes, assigning the Novokumak horizon to the 18th— 17th centuries BC. 

Cheek-pieces of Type II, decorated with ornamental compositions fully 
analogous to the art of the shaft graves at Mycenae, date to the 16th century BC. 
But some examples of this type also bear degenerate post-Mycenaean decoration, 
characteristic for the Carpathians and Danube region of the 16th 15th centuries 
BC. 

The similarity of detail of the most developed Eurasian Type III cheek-pieces 
with those of Mycenaean Type V provides a basis for synchronization. The 
cheek-pieces of Type V were found in tholos A at Kakovatos in a LH II context 
(dated to the 15th century BC; Palmer 1965: 171), in grave 7 at Dendra (LH 
IIIA, i.e., 14th century BC), in the 'House of the Shields' at Mycenae which was 
destroyed in LH IIIB, i.e., the 13th century BC, and, consequently, belonged to 
the 15th-14th century BC (Wace 1960: 42). Thus the development of the latest 
Type III, the disc-shaped cheek, is set no later than the 15th— 14th centuries BC 
and it also serves as a terminus ante quem for the cheek-pieces of Type II, which 
belong to the 16th century BC. The presence of 'Mycenaean' ornament on the 
disc-shaped cheek-pieces both in the Danube region and also on the Eurasian 
steppe as far as Kazakhstan, a well known phenomenon that includes other 
categories of material, provides evidence of the influence of the highly devel- 
oped Mycenaean civilization on the cultures of Europe, including the East (Suli- 
mirski 1970: 43-50, fig. 66; Beresanskaja 1971: 7, 8, Smirnov and Kuz'mina 
1977: 43-50), brought about by the course of trade items (Smith 1987). However, 
scholars of the Mycenaean civilization suggest that the Mycenaean dynasty came 



TRANSPORT 123 

to Greece from the north and it is in that direction that they seek a series of traits 
of Mycenaean material culture. The character of the chariot myth speaks for a 
central European origin for them. A specially important mythologem is the 
swan-horse, carrying the solar god across the sky (Ivanov 1969: 54); the 
possibility of a Mediterranean origin is excluded as this myth could only come 
out of northern Eurasia, where the flight of the swan signals the coming of 
spring. In Greek myth the chariots of the Amazons, pulled by swans, fly from the 
Rhipian mountains, i.e., the Urals. 

In pre -Mycenaean Greece there was no developed tradition of horse transport 
that was not stimulated by the invention of the horse-drawn chariot. In contrast, 
on the Eurasian steppe horsemanship was known from the 4th millennium BC to 
the 3rd millennium in a zone from the Dnieper to the Ural which defined the 
horse -using cultures, the formation of a cult of the horse (Kuz'mina 1977) and 
the tradition of sacrificing horses in burials, which only appeared in Greece with 
the arrival of the Mycenaean dynasty; the burial of two horses was uncovered in 
the dromos of a tholos at Marathon (Vanderpool 1959: 277-283). In contrast, in 
the steppe the ancient horse -raising tradition led to the development of horse- 
training. The intensive quest for the most effective methods of harnessing is 
evident from the wide variety and instability of variants of cheek-pieces in the 
steppe on sites of the Potapovka and Petrovka-Sintashta cultures. All of this 
points to the conclusion that the formation of cheek-pieces of Type I originated 
in the Eurasian steppe where the largest number of cheek-pieces are known, and 
the most archaic examples devoid of any decoration are known from the Aba- 
shevo and Catacomb-Multi-roller Ware cultures. On the border of the Middle 
and Late Helladic periods the chariot and paired horses with disc-shaped cheek- 
pieces from the steppe found their way into Greece. 

Cheek-pieces without tenons 

This group comprises cheek-pieces with a flat shield and a large central hole; 
usually there is a transverse side slot and a protruding triangular tab with small 
holes. Two subtypes are distinguished, la with one extension, lb with two exten- 
sions (Figs. 37, 38). 

Subtype la, variant 1, with round shield is present in burials of the Alakul' 
type in the Urals such as Alakul' (together with a pair of horses, Sal'nikov 1952: 
58, fig. 7.2); in northern Kazakhstan at Novonikol'skoe (AO 1977: 514, fig. on p. 
513) and Petrovka 2 (Zdanovich 1983: 63); and also on monuments of mixed 
types, e.g., the Kozhumberdy settlement at Tasty-Butak in western Kazakhstan 
(Sorokin 1966: table xxxviii, 22), in the Amangel'dy burial at Aydabul' in nor- 
thern Kazakhstan (Akishev 1959: 16, fig. 8); they are also known as chance finds 
in the Ukraine such as Diky Sad and Surskaya Zabora, or Voloshskoe (Sharafut- 
dinova 1968: 27, fig, 3,52; 1982: 139-140, fig. 55.7, 11, 56, 59, 67; Cheredniche- 
nko 1977: fig. 2.3). A variant of this type is a cheek-piece from a Timber-grave 
settlement of Polyani on the Severskiy Donets (AO 1975: 383; Berezanskaya 
1982: fig. 40.2). In contrast to the examples from Tasty-Butak and Vatina it has 
additional small slots on the triangular slat. 

Two cheek-pieces from the Timber-grave settlement of Kazangulovo 1 in Ba- 
shkiria bore oval and not round shields and these should be assigned to a special 
variant 1A 2 (Gorbunov and Obydennov 1975: 254-257, fig. 1.1,2). Cheek-pieces 



124 CHAPTER EIGHT 

of subtype 1A, as K. F. Smirnov (1961a: 60-63, fig. 9,1,11,1; type II in his 
classification) remarked, are close to Danubian ones mentioned by A. Mozsolics 
(1953: fig. 6.4-6; Bokonyi 1953: 11, fig. 2; Foltiny 1967: 28, fig. 6.7). Cheek- 
pieces from Vatina and an example from Toszeg belong to subtype IA (Htlttel 
1978: fig. 9; 1982: fig. 4); other examples from Toszeg belong to subtype IB. 

As far as the origin of shieldlike cheek-pieces without tenons is concerned, it 
is probable that they can be derived from shieldlike cheek-pieces with tenons of 
types III and IV with which they are related through the presence of a side slot 
through another tab (Kuz'mina 1980a: 15, fig. 1). Their evolution along both the 
Danube and across the Eurasian steppes occurred contemporaneously through 
the preservation of active contacts between the regions which is evident not only 
from the resemblance with respect to form and construction of the Hungarian 
and Timber-grave/ Andronovo cheek-pieces but also because they bear the same 
post-Mycenaean ornament — circular with points — decorating shields of two 
Toszeg examples (Hiittel 1982: fig. 3.5, 8) and fromNovonikol'skoe, a degraded 
Mycenaean wave from Kazangulovo and decayed post-Mycenaean rosette from 
Vatina. These Danubian analogies can help dating the Andronovo cheek-pieces. 
A. Mozsolics (1978) assigned all disc-like cheek-pieces to the Early Bronze Age 
in Hungary, which is defined by her as the 14th— 12th centuries BC without divi- 
ding them into classes and types; I. Bona (1975: 105-139) dates Toszeg and Va- 
tina to 1450-1200 BC. A date of the 15th-13th centuries BC seems most reliable 
and agrees with radiocarbon dates for the Novonikol'skoe burial that places it in 
the 14th century and for the Tasty-Butak settlement which is 1229± 80 BC. 

Grooved cheek-pieces (Figs. 37-39) 

Another class of cheek-pieces consists of cheek-pieces with protruding teeth- 
tenons on the edges (Figs. 37-39, 80: 35). They have a large central hole and 
separate tab. They are divided into two main types: 1) without a side slot, 2) with 
additional side slot. 

Type 1 is represented by an example from the early Petrovka-Alakul' settle- 
ment of Mirny 4 in the Urals (Chemyakin 1974: 55, fig. 2,6) where a cheek-piece 
of oval form was found with a central oval slot and separate shield. Another is a 
cheek-piece of rectangular form with a large rectangular hole and separate 
triangular tab with a small hole. Cheek-pieces of this variant are found in the 
Petrovka burials of Satan in central Kazakhstan along with a chariot (Tkachev 
1987: fig. 1.4) and Aksaiman in northern Kazakhstan together with a pair of 
draught horses (Zdanovich 1988: 78, fig. 30.22, 23). Two cheek-pieces of a well- 
developed variant with cross projections on the slat like those on the shieldlike 
cheek-pieces without tenons (Alakul', Kazangulovo, Toszeg) were also found 
(Zdanovich 1988: fig. 30.24, 25). 

Cheek-pieces of Type I are archaic in form and in construction and are 
comparable to disc-like cheek-pieces with tenons of subtype IIB, which 
establishes their lower data as the 16th century BC. Zigzag ornament on the 
shield of the Satan cheek-piece does not contradict this; it is analogous to the 
decoration on Type II cheek-pieces although it is known later as well. 

Type II is presented by classical rectangular grooved cheek-pieces with a side 
hole on a second slat. On the basis of the separate tags with small slots, the 
grooved cheek-pieces are divided into two subtypes: IIA with rectangular slat, 



TRANSPORT 125 

2B with triangular slat. One cheek-piece of Type II was found on the Kozhum- 
berdy settlement of Tasty-Butak; its slat was broken (Sorokin 1966: table 
xxxviii, 30). Subtype IIA cheek-pieces are found in central Kazakhstan on the 
settlement of Ikpen' in a complex with Fedorovo type pottery (Fig. 80: 35) and 
in the Ukraine on settlements of the Multi-roller Ware culture such as 
Podgorovka (Berezanskaya 1982: 127, fig. 40.3) and the early Timber-grave 
settlement of Osipovka (Belyaev 1980: 15), on the Severskiy Donets and Don, on 
settlements with mixed Multi-roller Ware and Timber-grave -Abashevo ceramics 
such as Kapitanovo, Prokazino, Shilovo, and on early Timber-grave sites along 
the Volga in burials both at Usatovo and Krasnopol'e, and on the settlements of 
Moechnoe Ozero and Tochka (I am grateful to I. B. Vasil'ev for this informa- 
tion). Cheek-pieces of subtype IIB have been discovered on the Donets Timber- 
grave sites such as Il'ichevo and along the Volga at settlements such as Ershovo 
and Guselka 2, and the kurgan at Komarovo which included a pair of chariot 
horses, one of which was bridled; in addition, grooved cheek-pieces are found in 
the Timber-grave complexes of Novomolchanovka, Cherebaevo, Stepovoe 
(Smirnov K. 1957a: 47-50, fig. 1.2; Cherednichenko 1968: 102-104, fig. on page 
103; 1970: 234, fig. 2.1,5; Bratchenko 1976: 151, fig. 72, iv,2; Pryakhin 1976: 
fig. 10, 11; Shapovalov 1976: 158, 160, fig. 4.1; Berezanskaya et al. 1986: 100). 

Zigzag ornament is found on some of the rectangular cheek-pieces (Usatovo, 
Komarovo, Tasty-Butak). It is characteristic of disk-like cheek-pieces with 
tenons (Mycenae, Staroyur'evo, Sarata-Monteoru, Ulmeni), and it is preserved 
on disk-like cheek-pieces without tenons (Novonikol'skoe). Another cheek-piece 
comes from the Timber-grave settlement of Chishma in Bashkiria; it is decorated 
along the edge by curved zigzags, and from above by concentric circles (AO 
1980: 144). This post-Mycenaean ornament is characteristic of disc-like cheek- 
pieces without tenons from Novonikol'skoe and Toszeg and provides an upper 
date for grooved cheek-pieces. 

In the peculiarities of construction (presence of tenons, central hole, trans- 
verse slots, and tag with small holes) Type II grooved cheek-pieces are closest to 
the most developed Type III disk-like cheek-pieces. Their decorative ornaments 
are also analogous, which aligns them to a single line of bridle development and, 
therefore, synchronizes them and sets their lower date to the middle of the 2nd 
millennium BC. K F. Smirnov (1961a: 52, 57) observed the similarity (in the 
method of bridle fastening) between grooved cheek- pieces and Hungarian cheek- 
pieces of another class, the horn cheek-pieces of Fiizesabony type with a large 
central hole, transverse hole and small holes on the end. They are assigned to the 
Middle Bronze Age, Fuzesabony-Toszeg-Ottomani horizon. According to A. 
Mozcolics, they date to 1400-1100 BC. A. Tocik (1959: 42-53, table II.3) noted 
a Fiizesabony type cheek-piece from a settlement of the Modarovskaya culture in 
Slovakia in which objects with Mycenaean ornaments are known. A. Mozcolics 
(1960: 133ff) attributed great significance to Anatolian influence. S. Foltiny 
(1967: 23, fig. 7.6) related the Fiizesabony type with a cheek-piece from Alaca 
Hiiytik from the Hittite layer of 1450-1300 BC and a cheek-piece from Megiddo 
of 1200 BC. He especially emphasized the finding of a bone disk analogous to a 
bronze disk from Kakovatos which was already compared (Hachman 1957: 174- 
175, fig. 7.10-12) and found together with a cheek-piece. R. Bochmer (1972: 
201) believed that the Alishar and Alaca Hiiytik cheek-pieces (of unspecified 



126 CHAPTER EIGHT 

date) and the Bogazkoy cheek-piece can be no earlier than 1450 BC. T. Kovacs 
(1969: 159-164) examined the horn cheek-pieces from the Carpathians and dated 
them on analogy with cheek-pieces from Asia Minor from Beycesultan and 
Alaca Httyiik to 1450-1300 BC. H. G. Htittel's (1982) skepticism about a 
connection between the Anatolian and Danube rodlike cheek-pieces is not well 
founded, but he does justly stress that their Asia Minor origin is not proved. T. 
Kovacs (1969: 164), I. Bona (1975: table 105) and B. Hansel (1968) lowered the 
lower date of the Hungarian sites of the Middle Bronze Age to the 16th- 15th 
centuries BC on the basis of objects with Mycenaean ornament. 

In order to specify when grooved cheek-pieces began to appear it is important 
to take into account the finding of a grooved cheek-piece with a bone plate with 
Mycenaean ornament on the early Timber-grave settlement of Il'ichevo (Shapo- 
valov 1976: 158, 159, fig. 5; Smirnov and Kuz'mina 1977: 47, fig. 12.6). On 
analogy with Mycenaean ornaments and a bone plate from the Vatina culture 
(Marinatos 1960: 90, fig. 202; Tasic 1973: 32, table IV.4) its lower date can be 
set to the end of the 16- 15th centuries BC. The setting of the lower date to the 
middle of the 2nd millennium BC is also secured by culturally mixed sites 
reflecting the interconnections of the Multi-roller Ware, Abashevo and early 
Timber-grave ceramics of the Pokrovskiy stage. 

The upper date of grooved cheek-pieces is established on the basis of the 
discovery of a grooved cheek-piece with a disk-like cheek-piece without a tenon 
on the Kozhumberdy settlement of Tasty-Butak; its age is defined as the 15th- 
13th centuries BC and by a radiocarbon date of 1229±80 BC. 

Type III is represented by the only cheek-piece from the Chelkar settlement 
in northern Kazakhstan (Orazbaev 1958: 275, tab. ix, 1), its tag is protruded and 
it has no holes. It probably reflects a degradation of the Type II cheek-piece. The 
absence of close analogies makes its dating impossible. A. M. Leskov (1970: 45, 
fig. 30.4) compared it with a cheek-piece from the settlement of Kirovskiy and 
assigned it to the 13th century BC. A cheek-piece from the settlement of 
Mosolovskiy along the Don is also unique: it has two holes on a rough tag 
(Pryakhin 1992: fig. 6). 

How Type I and II cheek-pieces were fastened is not entirely clear. On the 
basis of a grooved rectangular cheek-piece on a horse skull from the Komarovo 
kurgan, K F. Smirnov (1961a: 50-51, fig. 3) suggested that the horse was bridled 
with the help of nose and lip belts; a soft leather bit with a loop was passed 
through the central slot to which the rein was fastened; the cheek belt was 
fastened to the side slot, and small holes on the tag were for fastening with a 
nose belt; the tenons on the cheek-pieces were used as snaffles (Smirnov 1961a: 
63). (For a reconstruction of the fastening system of different cheek-pieces see 
Littauer 1969: fig. 7; Gorbunovand Obydennov 1979; Pryakhin 1992). 

There is no instance of a Type I cheek-piece found in situ in a Petrovka 
grave. The images of chariot horses with disk-like cheek-pieces on frescoes from 
the palace at Tiryns are important for reconstructing the bridle (Wiesner 1968: 
fig. 15, a, b). N. N. Cherednichenko (1976) supposes that bridle fastening was 
close to that described. 

We see then that two classes of cheek-pieces (shield and grooved) appeared 
over a wide zone of the Eurasian steppes from the Dnieper to northern Kazakh- 
stan in the Multi-roller Ware culture, Abashevo, early Timber-grave (Potapovka 



TRANSPORT 127 

type) and early Andronovo (Sintashta, Petrovka types) and developed further in 
the Timber-grave -Pokrovskiy, and Andronovo-Alakul' and Fedorovo complexes. 
The difference in form of both classes was preconditioned by peculiarities of the 
raw material: shieldlike cheek-pieces were made of a wide part of horn or from 
the epiphises of a cow's metatarsus (Balanbash, Kamenka), which results in a 
round, oval or segmented form. Grooved cheek-pieces were cut from a longitu- 
dinally split tubular bone, resulting in their rectangular or oval form, grooved in 
cross section. The principle of construction of the cheek-pieces is similar for 
both classes: they have protruding tenons and a large central slot; they evolved in 
the same direction, starting with a tag with small slots (rectangular as in variant 
A or triangular, variant B), then a transverse hole appears. Finally, they 
employed flat (without tenons) cheek-pieces with transverse holes. 

Two classes of cheek-pieces, rectangular and disk-like (according to J. H. 
Potratz, Type I were of bronze and not bone) were also used for harnessing 
chariots in the Near East (Potratz 1966: 103-1 16, fig. 45, 46; table 106-109,1 15). 
They are found in Gaza (Tel-El-Adzhul), Assur, Ras-Shamra, new chance finds 
of disk-like cheek-pieces are housed in the Jerusalem and Metropolitan museums 
(Littauer and Crouwel 1982: table xvii; 1986b, table xlii) and also in Egypt in 
Tel-El-Amarna and in Greece at Mycenae (Crouwel 1981: table 1.2-4). Unfortu- 
nately, stratigraphic contexts are not clear for many of them. Typologically, the 
earliest rectangular and disk-like cheek-pieces from Gaza, according to F. Petrie, 
are connected with the Hyksos and they date from the 17th century BC, a date 
which is accepted by J. H. Potratz (1966: 109, 110, 115 116) or the 16th-15th 
centuries BC according to K. Sheffer: a disc-like cheek-piece from Tel-El- 
Amarna (without tenons) is dated 1413-1377 BC, a rectangular one to 1375-1350 
BC. An image of a disk-like cheek-piece has been found in Abydos in the temple 
of Ramesses II (13th century BC) (Frene 1965: 105). A rectangular cheek-piece 
from Gezer has been assigned to 1300-1100 BC. H. G Htittel (1978: 10) dates 
all Near Eastern cheek-pieces to the 16th 13th centuries BC. The terminus ante 
quem for a disk-like cheek-piece from Ras-Shamra is the destruction of the city 
at the end of the 13th century BC. M. Littauer and J. Crouwel (1979: 208; 1982: 
178; 1986: 165) assign the wheel-like metal cheek-pieces of Gaza, Amarna, 
Mycenae and others to the second half of the 2nd millennium BC. 

The correlation of the different types of bone and horn cheek-pieces across 
the Eurasian steppes, in the Danube region and Mycenae, and the bronze cheek- 
pieces with tenons of the Near East has been discussed many times. However, 
hypotheses concerning the origin of the cheek-piece are diametrically opposed. 
A. M. Leskov (1964: 302), K. F. Smirnov (1961a: 57, 71-2; Smirnov and Kuz'- 
mina 1977: 43), E. E. Kuz'mina (1971b: 92-3; 1977; 1980a: 18), N. N. Cheredni- 
chenko (1976: 147-148), and V. B. Kovalevskaya (1976: 60-61) suggest that 
Near Eastern and Mycenaean cheek-pieces derive from steppe cheek-pieces. 
According to A. M. Leskov they come from Central Asia; K. F. Smirnov, E. E. 
Kuz'mina and V. B. Kovalevskaya suggest south Russia; according to N. N. 
Cherednichenko they derive from eastern Europe and then passed to Mycenae 
and then on to the Near East. A. Mozsolics (1960: 125) admits that the proto- 
types of cheek-pieces in the Danube region were borrowed from the Pontic 
steppes, but the main importance is attributed to Anatolian influence. G Potratz 
(1941) considered the Danubian cheek-pieces to be the prototypes of the Near 



128 CHAPTER EIGHT 

Eastern cheek-pieces. H. G. Hiittel (1978) notes the similarity of the steppe, 
Danubian and Mycenaean disk-like cheek-pieces but insists on the independent 
development of European and Near Eastern cheek-pieces of different classes and 
he rejects diffusion. In contrast, B. A. Latynin (1965: 203-204), V. A. Safronov 
(1966: 189) and A. Oancea (1976: 75) believe that the disc-like cheek-pieces 
appeared along the Danube and the steppes from the Mediterranean; according to 
B. A. Latynin they derived from Anatolia, according to V. A. Safronov and A. 
Oancea from Mycenae. M. Littauer (1969: 298), stressing some differences in 
construction, asserted that bone cheek-pieces of the steppes are imitations of 
metal ones from the Near East. Later she and J. Crouwel (1979) stressed that 
Near Eastern cheek-pieces developed independently. 

Hypotheses about a Near Eastern origin for the steppe cheek-pieces can 
probably be rejected on the basis of the construction of the bone cheek-pieces, 
their form, size, tenons and central hole with a reinforcing bevel, which were 
originally dictated by the specific character of the initial raw material; then the 
formal peculiarities were repeated in metal. An expressly Mycenaean bronze 
cheek-piece was found in grave 81: its construction included inserted tenons 
which reproduced an important improvement of bone cheek-pieces, but one that 
was no longer necessary in the more solid bronze ones. Cheek-pieces from 
Gezer, Amarna, Mycenae and Cyprus have a bronze twisted bit, imitating a 
softer material. Thus, the first cheek-pieces were made of bone and horn with a 
strap bit; bronze cheek-pieces originated on this basis. 

An Anatolian origin for the steppe chariot and cheek-pieces is also improb- 
able. At Sharkishly in central Anatolia only disc-like cheek-pieces without 
tenons are known; rectangular and disc-like cheek-pieces were not used but 
rather rodlike cheek-pieces of horn which were not found in the Eurasian steppe 
(Hiittel 1978: 10, fig. 3). Rodlike cheek-pieces of horn, round in section, can be 
regarded a Hittite form. They were found at Alaca Huytik in a layer of c 1450- 
1300 BC, at Beycesultan about 1450-1300 BC, in Bogazkoy and Alishar; there is 
one find from Megiddo, Palestine. In terms of construction and ornament such 
cheek-pieces are closely analogous to Hungarian cheek-pieces of the Fuzesabony 
type (Foltiny 1967: 23). In the Carpathians and Danube basin two main types of 
horn cheek-pieces were distributed, Fuzesabony (with one central hole) and 
Toszeg (with two central holes). They coexist in a single complex, but some 
Toszeg cheek-pieces are decorated with the Mycenaean wave ornament and they 
probably originated somewhat earlier, in the 16th century BC. The similarity 
between the Hittite cheek-pieces and the Carpathian-Danubian ones reflects 
without doubt active cultural connections between these two regions. The 
peculiarity of this form, which was not found in other regions, probably suggests 
a genetic kinship between two groups of horse-breeding populations. If this 
assumption is correct, it agrees well with the linguistic hypothesis of a Hittite 
migration from the Carpathian-Danube region. 

The cartographical distribution of the earliest three main classes of cheek- 
pieces in the Old World in the middle of the 2nd millennium BC — disc-like with 
tenons, rectangular cheek-pieces with tenons and rodlike cheek-pieces of horn — 
shows that disc-like cheek-pieces were used in the steppes, along the Danube, in 
Mycenae and in the Near East, except for the Hittite kingdom; the rodlike ones 
were found along the Danube and in the Hittite realm; rectangular with tenons 



TRANSPORT 129 

were distributed in the steppes, Mycenae and in the Near East. The synchronous 
character of the two latter classes is indicated by the similar construction of 
bridle and the ornaments mentioned by K. F. Smirnov. Their predominance in 
different territories indicates the specific character of the traditions of two dif- 
ferent groups of horse-raising peoples. As was mentioned above, rodlike cheek- 
pieces are specific to the culture of the Hittites, as well as peoples of the Carpa- 
thians and the Danube regions, where many linguists see peoples related to the 
Hittites such as the Thracians, Phrygians, etc. Disc-like horn and bone cheek- 
pieces with tenons were found in the culture of the Achaeans, the Indo-European 
people of Mycenaean Greece, in the steppes and among supposedly Indo-Europ- 
ean populations of the Carpathian-Danube zone. Rectangular cheek-pieces (bone 
and bronze) are characteristic of the steppes and the Near East, except Anatolia. 
Rectangular cheek-pieces with tenons survive in the Near East till the Assyrian 
period. The invention of a second small slot that provided a firmer fixing of the 
cheek-pieces and the use of a forked cheek belt improved their form. Disc-like 
cheek-pieces with tenons were used in Luristan till the Iron Age (Potratz 1966). 
Disc-like cheek-pieces without tenons are spread in other regions; their lower 
date is established according to findings at Amarna of 1413-1377 BC. 

In construction, the earliest Near Eastern bronze cheek-pieces with 
rectangular plate, with a central hole and small slots above the plate, in terms of 
their form and the presence of a bevel around the central hole, are closest to a 
cheek-piece from Mirny; in terms of the construction with a transverse hole they 
are close to the grooved cheek-pieces with a rectangular plate known from the 
steppe. The bronze disc-like early eastern cheek-pieces with tenons and an 
additional side slot for fastening the cheek belt resemble most of all the steppe 
cheek-pieces of Type 3 (Staroyur'evo). If these comparisons are valid, they are 
of principal importance, suggesting the natural habitat from which the prototypes 
of cheek-pieces of both classes could appear in the Near East. As bone disc-like 
cheek-pieces with tenons combine with rectangular cheek-pieces with tenons 
only in the Eurasian steppe from the Dnieper to central Kazakhstan, we can 
propose that this zone was the initial territory whence cheek-pieces originated in 
the Old World. If this conclusion is true, it is very important in resolving the 
problem of Indo-Iranians origins. 

Rodlike cheek-pieces (Figs. 39: 4-7; 42: 10) 

Type 1 is a cheek-piece which is three-holed, slightly curved, round in section, 
with an oval slot in the center and with one or two small transverse round slots. 
A fragment of this type was found on the settlement of Atasu in house 4 in a 
com-plex with ceramics with applied-roller ornament (Margulan et al. 1966: 
table xxiv, 7). Cheek-pieces of the Sabatinovka stage of the Timber-grave culture 
from the settlement of Kirovo in the Crimea, and Voinskaya Greblya, Skelya, 
Il'ichevka in the Ukraine (Terenozhkin 1965: 70-71; Sharafutdinova 1968: 16- 
34; 1982: 138-139, fig. 55, 4-5; Leskov 1970: 45-46, fig. 30.5; Shapovalov 1976: 
164-5, 170, fig. 4,2; Berezanskaya 1982: 40; Berezanskaya et al. 1986: fig. 29.4, 
7,10). A metal cheek-piece of this type was found on the settlement of Kuzeli- 
Gyr in Khorezm (Terenozhkin 1958: 34-39, fig. 1; Kuz'mina 1966: 59-60, table 
xv, 41). The type is spread over a vast territory of Eurasia, on the sites of 
Worschach and §ant-Rodna Noua that are assigned to the Bronze D and Hallstatt 



130 CHAPTER EIGHT 

A periods, i.e. to the 12th 1 1th centuries BC according to H. Mtiller-Karpe and it 
is close to Hungarian cheek-pieces (Rusu 1960: 161-180; Pitioni 1954: 475; 
Smirnov K. 1961a: 63-65) dated by A. Mozsolics to c 1100 BC, according to 
other scholars to the 13th 11th centuries BC (Smirnova 1970: 106-110), 
indicated by a radiocarbon date from Il'ichevka of 1360±40 BC. Cheek-pieces of 
this type are also known in western Siberia, on the post-Andronovo settlement of 
Elovka (Matyushchenko and Lozhnikova 1966: 188, fig. 5). 

Type II is represented by cheek-pieces, oval in section, with three holes in 
one plane (the central one is oval), sometimes with a noticeable widening at the 
holes. It is known from the settlement of Yazevo 1 on the Tobol (Potemkina 
1979: fig. 6.4; 1985: 289, fig. 17.1, 108, 237). Cheek-pieces with small trans- 
verse holes are found at Shortandy-Bulak in central Kazakhstan (Margulan 1979: 
fig. 161.16) in a complex with ceramics with applied-roller. S. Ya. Zdanovich 
(1979: 12) mentions the finding of this type in the Sargary complex in northern 
Kazakhstan. Two straight cheek-pieces of this type are found in the settlement of 
Yukalekulovo of the Cherkaskul' (Mezhovka) culture in the Trans Urals 
(Obydennov 1986: fig. 3b, 38, 29). They are assigned to Type V according to K. 
F. Smirnov (1961: fig. 12). This type is also characteristic of Late Timber-grave 
sites of the Volga (Zhirnokleevka), of the Maklasheevka stage of the Pre-Kazan' 
and the late Sabatinovka cultures, Belozerka and Chernolessk in the Ukraine 
(Dereivka, Usatovo, Subbotovo) (Smirnov K 1961a: 66; Khalikov 1969: fig. 55, 
no. 163; Leskov 1971: fig. 2, no. 18,19,4; Sharafutdinova 1982: 140, fig. 55, 8- 
10; Berezanskaya 1982: fig. 40.5). The type is spread over a wide territory, 
including post-Mycenaean Greece, where according to S. Foltiny (1967: 30, fig. 
13) it was introduced by steppe nomads, Hungary where, according to A. 
Mozsolics (1960: 90-93), it is dated to the pre -Scythian period and has an 
Eastern origin, the northern Caucasus where it is known on the monuments of 
the Koban culture, and western Siberia, Bol'shoy Log, Elovka, Irmen (Gryaznov 
1956b: 74; Gening, Gusentsova et al. 1970: 46-47, fig. 1, 113; Matyushchenko 
1974: 58-60, fig. 13.4, 30.7, 45.3) where it is appears in the Karasuk period and 
survived until early Scythian times. The Eurasian three-holed cheek-pieces of 
this type were the prototypes of the three -holed tubular cheek-pieces of the early 
Scythian period that appeared in the 8th century BC and it provides the terminus 
ante quern for this type (lessen 1953: 109, 1 10, fig. 24; 1954: fig. 7). 

Of great importance for the chronology of the sites of the Alekseevka type is 
the discovery of a curved cheek-piece (Fig. 42: 10) with three holes and two 
small transverse slots that was found in the settlement of Kent in central 
Kazakhstan (Varfolomeev 1988: 87, fig. 5.20). It is analogous to an example 
from the Sabatinovka settlement of Chikalovka in the Ukraine (Sharafutdinova 
1970) which dates to the 13th 1 1th centuries BC in the Danube region. The 
complex of the Kent settlement (Fig. 42) includes bone objects with circular 
ornament of the Noua-Sabatinovka culture (Fig. 87), a large set of bronze objects 
characteristic of the hoards of the Final Bronze Age and imported ceramics of 
Namazga VI type (Varfolomeev 1987: fig. 6; 1988: fig. 3,4,6; 1991: 15, 20). It 
helps us synchronize the European and Central Asian lines of development. 

A cast model for the production of a bronze cheek-piece of type II was found 
in Central Asia on the settlement of Dal'verzin (Kuz'mina 1966: 60, table 
15.40). Similar bronze cheek-pieces were found in Iran, in Giyan 1, Sialk VH3, 



TRANSPORT 131 

Hasanlu (Fig. 106; Conteneau and Ghirshman 1935: table v. 6; Ghirshman 1939: 
tables lvi, lvii, 1963: fig. 338), and, taking into account an origin of this type in 
the Eurasian steppes, it is in good accord with the hypothesis originally proposed 
by R. Ghirshman about the arrival of Iranian-horsemen at Sialk VIB from the 
Eurasian steppes. The date of metal cheek-pieces of this type is defined by the 
Giyan and Sialk complexes to the beginning of the 1st millennium BC which 
provides a terminus ante quern for the appearance of bone cheek-pieces. The 
upper date of the latter is defined by their appearance in Assyria on an Assyrian 
relief of Assurbanipal (668-629 BC). H. Potratz (1966: 28, 33) explains this by 
contacts with the Medes and Scythians. 

In light of the discovery at Kent of numerous objects of horse gear of pre- 
Scythian date (Varfolomeev 1987: fig. 6.5-7; 1988: fig. 4.10-13, 5), it is natural 
that we also find bronze three -holed cheek-pieces, a stirrup-like bit, buckles, 
perforated bridle ornaments of early Scythian type with analogies at Arzhan in a 
Final Bronze Age burial at Izmaylovka in eastern Kazakhstan in a complex with 
vessels of Begazy and Dongal types (Fig. 107; Archaeological monuments 1987: 
figs. 32, 33). They complete almost a millennium of development of horse gear 
in the Andronovo culture and open a new, Scythian, page in the history of the 
steppe. Three -hole cheek-pieces with a stirrup-like bit of the Arzhan type are 
found to the north of India in Swat, in the Timargarha cemetery (Lesnik 1971). 

Frontlet (Fig. 39: 2) 

Among the horse-gear of the late Andronovo period, one should mention a bone 
frontlet from house 4 at the settlement of Atasu, found in the same complex as 
applied-roller pottery and a cheek-piece (Margulan et al. 1966: 215, table xxiv, 
8). A similar object was found in the Sabatinovka II layer of the settlement of 
Il'ichevka on the Donets which also had ceramics with applied-roller ornament 
and a three -hole cheek-piece of the Boriyash type, analogous to the Atasu one, 
which defines the dates of both complexes and affirms their synchronization 
(Shapovalov 1976: 165, fig. 7.19). The dating of the frontlets is specified by the 
discovery of a frontlet with the image of a sphinx in the palace of Ashurnasirpal 
(1050-1032 BC) at Nimrud (Foltiny 1967: fig. 4). 

Thus, there is a continuous development of the bridle from the 17th to the 9th 
century BC in the Andronovo culture. Cheek-pieces of classes I and II of the 
17th 13th centuries BC were used for chariot harnessing, which is witnessed by 
their discovery in situ on the skull of harnessed horses and from their depiction 
on works of art in other places of the Old World. Class III, Type II cheek-pieces 
were probably used for horse riding, judging by their genetic connection with the 
early Scythian ones. Their appearance at the end of the 2nd millennium BC 
shows the time when horse riding spread across the steppes. 

The tactics of chariot warfare and horse riding 

Chariot warfare dominated in the 16th— 12th centuries BC over a vast territory of 
the Old World — across the Eurasian steppes, along the Danube, in Mycenaean 
Greece, the Near East, Egypt and later in China. 

According to linguistic data wheeled transport and horse-raising were 
especially wide-spread among the Indo-Iranians. Apart from common Indo- 



132 CHAPTER EIGHT 

European 'transport' terms the Indo-Iranian languages established a developed 
horse -raising vocabulary that included terms for the color of the horse, its age, 
parts of the body, fodder, etc.; there were also common words for 'chariot', 
'shaft', 'bridle', 'saddle-girth', and many names of harness parts (Abaev 1949: 
54; Bailey 1955: 1-14; 1957), as well as the names of rituals connected with the 
horse (Ivanov 1969a; 1969b; 1974; Dumonte 1927; Kane 1946; Renou 1954). 
The expression 'managing horses' is applied to the highest official; the term 
raOaestar- 'standing on a chariot' denotes a representative of the privileged 
warrior caste with special insignias (gold, red color and weapon), and a special 
role in society (Leroy 1946; 1957; Brandenstein 1962; Christensen 1934: 136; 
Dumezil 1930: 109-130; Benveniste 1932; 1938; Grantovsky 1960; D'yakonov 
1961: 60-62; HTP, 143-147; Bongard-Levin and II' in 1969: 162-176; Gafurov 
1972: 54-57; Kuz'mina 1975: 291-292; 1977). 

Where were chariots, that played such a revolutionary role in the military art 
of the Old World, invented? M. Littauer and J. Crouwel (1996: 938) repeat their 
old point of view that the chariots in the steppes were imitations inspired by the 
Near East. This conclusion is based on two erroneous premises. The first, follow- 
ing the theories of M. Gimbutas and D. Anthony, is that they believe that there 
were mounted warriors in the steppe since the Neolithic period. Secondly, fol- 
lowing E. Izbitser (1993), they believe that two-wheeled vehicles were unknown 
in the steppes in the 3rd millennium BC. However, two-wheeled carts have been 
found in Pit -grave and Catacomb burials of Storozhevaya Mogila, Pervokonstan- 
tinovka kurgan 1/8, Mar'evka kurgan 11/27, Lola, etc.; clay models are known, 
e.g., in Tri Brata kurgan (Novozhenov 1994: 133, 140) and in a collection of V. 
A. Safronov. Two-wheeled carts with disc or so-called crossbar wheels are found 
on petroglyphs in Kazakhstan (Novozhenov 1994: 89, 91, fig. 51). 

In the cemetery of Novonikolaevka of the Mikhaylovka culture a two- 
wheeled cart with crossbar wheels was painted in ocher on the stone roof of a 
grave, if I interpret this image correctly (Rassamakhin 1999: 47, fig. 3.56). Such 
light wheels are depicted on seals in Anatolia, on a cylinder in Hissar III in Iran, 
and on a silver cup from Afghanistan (in the Louvre and in the royal cemetery at 
Gonur; Fig 96: 1-3, 10). They are viewed as transitional to the spoked- wheel 
(Littauer and Crouwel 1977: fig. 21, Amiet 1989: 61, fig. 6; Dubova 2004: 277, 
figs. 36-39). 

A Near Eastern origin for the steppe chariots has also been asserted by M. V. 
Gorelik (1985). But he failed to distinguish between the chariot and those freight 
and road vehicles, the heavy military wagons that were harnessed to equids other 
than horses. This led him to erroneous conclusions on the time and place of the 
origin of the chariot in the Old World. In his table 2 chariots are represented only 
on images of the second half of the 2nd millennium BC. The date of chariots 
depicted on Syrian seals (table 3) was established on analogy with two samples 
from the Nuzi archive which dated to the second half of the 15th century BC. M. 
Littauer and M. Gorelik's hypothesis was supported by V. V. Trifonov (1996) 
who alluded to images of two-wheeled carts drawn by equids on Syrian- 
Palestinian seals. The latter are dated to the 18th— 16th centuries BC (Porada 
1979; Moorey 1986), but these images are too schematic. 

New evidence for the history of transport comes from excavations in Syria at 
Tell-Brak which dates to the 4th-2nd millennia BC and as ancient Nagar was 



TRANSPORT 133 

the capital of the Hurrian dynasty at the end of the 3rd millennium BC, and then 
an important trade center of the Mitanni state in the 16th century BC (Oates J. et 
al. 1997: 141-154). Indo-Aryan horse breeders were among the Hurrian popul- 
ation of this city which is seen in the Indo-Iranian horse -raising terms in the 
Hittite treatise of the Mitannian Kikkuli, Indo-Iranian names of gods and rulers 
in documents from Nuzi, the Amarna archive in Egypt, and in new texts and 
seals (Fig. 96: 8) from Tell-Brak which bear the names of the kings Artasumarra 
and Tusratta. Indo-Aryan dominance in the Hurrian world is usually explained 
by their role in the spread of horse-raising and training skills, light military horse 
drawn chariots, the compound bow and the creation of the marianni; the 15th 
century BC Mitanni played an important role in the Ancient East. Aleppo and 
Alalakh submitted to Parattarna, his son Sausattar controlled Cilicia and Assur, 
his grandson Artatama gave his daughter in marriage to Pharaoh Tutmos IV, and 
in the 14th century BC Tusratta's daughter became the wife of Akhnaton. In her 
rich dowry there were chariots with gold decorations, six thousand arrows and 
iron objects. In 1332 BC the Hittite king Suppiluliumas sheltered Sattiwaza, the 
son of the assassinated Tusratta, who then married his daughter. Suppiluliumas 
made a treaty with him sealed by the names of Hittite and Hurrian gods; at the 
end of the treaty the Indo-Iranian gods Mitra, Varuna, Indra, and Nasatyas 
(Mitrasil, Uruwanassil, Indara and Nasatianna) are mentioned. He also helped 
the fugitive regain his throne (Jankowska 1982). The Mitanni kingdom was 
destroyed by the Assyrian King Adad-Nerari (1305-1274 BC) during the rule of 
Vasasatta. Nagar and other cities were destroyed. All this happened about 1300 
BC according to dendrochronological dates. 

There has been a long discussion about the role of the Indo- Aryans in horse- 
raising and the spread of the chariot across the Near East. New material from 
Tell-Brak helps to resolve this problem. In layers of the 3rd/beginning of the 2nd 
millennium BC archaeologists discovered a hostel for traveling merchants, 
donkey burials, 102 cart models, 191 equid figurines and numerous images on 
seals analyzed by J. Oates (Oates 2001). 

It is known from written sources and seal images that carts belonged to gods 
and kings and they were used for ritual and cult purposes. Four-wheeled vehicles 
harnessed to two or four animals with the help of a yoke and pole, both with 
open and closed frame, dominate among the models from Tell-Brak. Seventy 
percent of the two-wheeled vehicles belong to the end of the 3rd - beginning of 
the 2nd millennium BC. They are represented by two types: 1) low maneuver- 
able carts with curved pole and a platform with a seat; 2) lighter and more 
maneuverable carts in which the driver stands or sits right above the axle, which 
is sometimes shifted to the end of the platform. These are prototypes of original 
chariots. 

All 271 wheel models have a protruding hub, 269 wheel models are tri- 
partite, sometimes they have tyres; one wheel is of the light type, with slots 
(cross-bar wheel). And only one wheel with an image of spokes, belonging 
probably to the Ur III period, provides early evidence of this type which was 
known in Mesopotamia and Anatolia from images on seals that date to the 
beginning of the 2nd millennium BC (Oates 2001). Thus, in the pre-Mitanni 
period, light two-wheeled carts were known from Tell-Brak but the actual form 
of the chariot had not yet appeared. 



134 CHAPTER EIGHT 

The donkey (ansu), onager ('desert donkey') and mule (kimga) are men- 
tioned in written texts. The donkey was the main transport animal. The skeleton 
of a tall donkey with a bronze bit assigned to the 3rd millennium BC was found 
on the settlement (Clutton -Brock et al. 1993: 209-221). Donkey skeletons are 
found in grave H 70 in Halava, they also belong to the end of 3rd millennium 
BC. Some figurines from Tell-Brak depict these large draught donkeys. Onagers 
were spread in Iran and Syria; from the Neolithic period they were the main 
game for meat; judging by a hymn to King Shulga they were hunted with nets. 
Onagers are very wild and cannot be domesticated. 

Mules are a hybrid of a donkey and a female onager. Brak-Nagar was a 
famous center of their breeding and trade. Texts from Ebla say that the mule was 
ten times more expensive than the donkey and it was considered a better animal 
for harnessing, especially for four-wheeled and ceremonial vehicles. Mules 
harnessed by a nose ring as bovids, by a nose belt or a muzzle-kaptzug which 
prevented them from biting, can be seen on the majority of models and cart 
images (Fig. 95: 3; 96: 4, 6, 7). 

As far as the horse is concerned, only donkeys and mules are mentioned in 
the documents from Tell-Brak. In Mesopotamia the horse zizi or ansu kur-ra 
(donkey of the mountains) is very rarely mentioned from the Ur III period. But a 
text of the Old Babylonian period recommends that one ride a donkey, not a 
horse. Among the vast osteological material from the ancient Near East horse 
bones are supposedly found only in Tell-Leilan; the identification of bones from 
Tell-Selenkaya as horse, made by S. Bokonyi, has been disputed (Boessneck and 
von den Drisch 1977). 

J. Oates considers that only four out of the 191 equid figurines depict horses 
as well as a figurine from Tel-Taya, and admits that even the four might be large 
Syrian donkeys. Greater doubts are connected with the extremely rare images 
that have been interpreted as riders (Fig. 96: 4-6,9; Moorey 1970). 

In light of all this, the animals harnessed to two-wheeled carts by nose rings 
and kaptzugs, depicted on seals including the one from Kill-Tepe, should be 
interpreted as mules and not horses (Littauer and Crouwel 1979: fig. 28, 29, 32). 
Just a few figurines from Kill-Tepe probably depict horses (Ozgiic 1950: table 
xxxiii, 1.2). 

Thus, although the horse may have been known in the ancient Near East in 
the 3rd millennium BC, it did not play any role. Large donkeys and mules 
remained the draught animals at the beginning of the 2nd millennium BC as 
well. Reliable evidence indicating the use of the horse is confined to harnessing 
to the original light chariots with spoked wheels and these are dated before the 
18th 17th centuries BC. The extreme rareness of the horse could hardly stimu- 
late the invention of horse-drawn chariots in the Near East. The fact that horse- 
raising terms and training skills were borrowed by the Mitanni from an existent 
Indo-Iranian tradition already in the middle of the 2nd millennium BC reinforces 
our belief that the Indo-Aryans played an outstanding role in spreading the 
horse-drawn chariot in the ancient Near East as well as in India. 

The Eurasian steppe was the homeland of horse-raising in the Old World. 
Horse bones are numerous on sites of the 3rd / beginning of the 2nd millennium 
BC. Ritual horse burials and their images on petroglyphs and in plastic art are 
known, indicating the creation of a horse cult (Kuz'mina 1977; 2000). Thus, both 



TRANSPORT 135 

prerequisites for the invention of the chariot were present in the steppes: the two- 
wheeled cart and horse, and a military situation that stimulated the introduction 
of new battle tactics. It supports the hypothesis of an independent invention of 
the chariot in the south Russian steppes. 

Near Eastern texts indicate that before the middle of the 2nd millennium BC 
a group of Indo-Aryans penetrated to the north of Mesopotamia and the southern 
part of the Armenian plateau (Ivanov 1968; Brandenstein 1948; Thieme 1960; 
Hauschild 1962; Mayrhofer 1966). They took a leading position in Mitanni and 
neighboring principalities which is seen in the names of Indo-Iranian deities, as 
well as Indo-Iranian personal names, including representatives of the ruling 
dynasty and nobility. The role of the horse-drawn chariot in this society is 
affirmed by the title of the ruler, literally 'managing horses', and names of rulers 
according to such etymologies as Tusratta (=*Tvisratha) 'having rushing 
chariots', Abiratta (=*Abhiratha) 'facing chariots', Bardasva (*Varddhasva, son 
of *Vrddhasva 'possessing large horses') (Burrow 1976: 30; cf. Elizarenkova 
1989: 429, 430). The term marianni, denoting nobility, from Indo-Iranian 
*marya- '(young) man', indicates the existence of a privileged caste of warriors, 
although I. M. D'yakonov rejected this etymology. In Mitanni, Assyria, and the 
Hittite kingdom special treatises were written on the training of chariot horses. 
The earliest is the manual of Kikkuli the Mitanni (14th century BC). It was writ- 
ten in the Hurrian language and then translated into Hittite, but the entire horse- 
raising terminology in it is Indo-Aryan (names for the colors of horses, fodder, 
race track, turnings on the track, chariot driver). These Indo-Aryan terms are 
repeated later in Assyrian treatises (Ebeling 1951; Salonen 1955-56; Kammenhu- 
ber 1961). According to W. Nagel (1966: 36) in the 16th-15th centuries BC a 
koine was formed in military art in the Near East, under the influence of an Indo- 
Iranian military aristocracy battle chariots and trained horses quickly spread 
from Mitanni. Horses were extremely expensive (in Nuzi a horse was seven 
times more expensive than a bull and ten times more expensive than a donkey). 
They were brought from afar, from Mitanni and northern regions; they were kept 
in the king's stables; they were sometimes presented to nearby rulers. The Hittite 
king Hattusilis III requested the Babylonian king Kadashman-Enlil II to "send 
me horses, high-bred horses and tall stallions". In the third quarter of the 2nd 
millennium BC images of the horse-drawn chariot are often met in the monu- 
mental and applied art of Egypt and the Near East. It means that horse -raising 
and horse-training skills for the chariot were brought to the Near East by Indo- 
Aryans and that they were decisive in the change of battle tactics and in the 
development of chariot warfare. 

A. Kammenhuber (1968) tried to disprove this thesis arguing that the horse 
was known in the ancient Near East before the arrival of the Indo-Aryans and 
also contested a number of etymologies. She was supported by I. M. D'yakonov 
(1970; 1988), but M. Mayrhofer (1974) rejected their main arguments. P. 
Moorey (1970) showed that although horse raising was known long before the 
arrival of the Indo-Aryans in the Near East (mainly in regions of Asia Minor) 
harnessing techniques were ineffective and the horse could not be used in battle. 
There are only a few passing remarks about Near Eastern peoples knowing the 
horse before the arrival of the Indo-Aryans (see Kuz'mina 1977). Most 
importantly, in spite of massive excavations there are practically no bones of 



136 CHAPTER EIGHT 

domestic horse in the osteological collections from Near Eastern settlements 
until the Hittite period (Boessneck and von den Driesch 1975: 209). We have 
only individual findings of horse remains in Asia Minor, e.g., Norsun-Tepe. 
Some suggest that these horses were domestic (Bokonyi 1991: 123-130) but 
others argue they were wild (Boessneck and von den Driesch 1976: 81-87). 

Indo-Aryans introduced chariot -war fare tactics in Mitanni; Nuzi texts indic- 
ate a standard set of chariot equipment (Zaccagnini 1977: 21-38) consisting of a 
bow, two quivers with 30 arrows in each, a spear, a leather shield, reins, whip, 
two horse-cloths, leather coats of mail (sometimes with bronze) for horses and 
defense equipment for warriors, for the arms and gurpisi (lit. 'neck cover') — a 
hood standing like a comb with parts reaching the neck and shoulders (see 
Chapter 7). It was made of leather sewn with tendons, sometimes supplemented 
with bronze plates. The hood had 140-190 plates, the sleeve plates were smaller. 

Socketed arrows and a flexible composite bow were innovations that the 
Indo-Aryans introduced to the Near East. The king of Mitanni, Tusratta, sent 
bows to Amenhotep III as a gift, and he presented three quivers with 90 arrows 
to Amenhotep IV. 

Unfortunately, excavations in Mitanni have not uncovered this armament set 
which eludes archaeological verification. It is worth mentioning that the hood, 
composite bow, socketed arrows and spear are typically Andronovo equipment; 
moreover, socketed arrows are not found in the Near East where hafted types do- 
minated. Socketed arrows were invented in the Eurasian steppes in the first quar- 
ter of the 2nd millennium BC. They were in warriors' quivers of the Sintashta- 
Petrovka period; in the Final Bronze Age they were introduced by pastoral tribes 
into the farming oases of south Central Asia, and saw their final development in 
the Saka and Scythian cultures and were again introduced to the Near East 
during their Asian campaigns (Kuz'mina 1966: 33-37; Medvedskaya 1982). 

The horse-drawn chariot and similar set of armament found in Mitanni is 
reconstructed for the Aryans of India. There are fifteen synonyms to denote the 
'horse' in Vedic language; more than twenty words denote 'way', 'movement'; 
the wheel and chariot make up an entire semantic field, symbolizing truth, law 
and cyclic world order (Elizarenkova 1982: 28, 41, 43, 44; 1999; Toporov 1981: 
147-151). The god Tvastar, the Creator, made the chariot with his own hands. In 
Iranian mythology one of the seven karsvars, the northern homeland, was called 
X v anira9a, the country with 'skillfully made chariots' (P'yankov 1979; 2000). 
Both in the Rigveda and the Avesta the chariot is the attribute of a deity. Four 
white horses draw the gods' chariots (Dhalla 1922: 183). The Indo-Aryans have 
a magic chariot Puspaka; in the Nart tales there is the Barsag wheel and a self- 
flying chariot. A special Rigveda hymn (6.47.26-28) is dedicated to the divine 
chariot. A particular spell was pronounced when the king mounted his chariot, 
"Oh, divine chariot, accept sacrificial libations!" "Let standing on you conquer 
what is to be conquered!" {Atharvaveda 6. 125). 

Vedic sources contain evidence about the warrior-charioteers. According to 
the Aitareya-Brahmana (7.19) the symbols of the ksatriya are "chariot, armor, 
bow and arrows"; according to the Satapatha-Brahmana (5.3.5) "the bow is the 
ksatriya's weapon". Bronze knives, awls, stone axes, maces and grindstones are 
mentioned in the texts (Rau 1973; 1983). 



TRANSPORT 137 

Vedic data are confirmed by Avestan (Dhalla 1922: 131-133); unfortunately, 
some texts have been corrupted during oral transmission. The gods Mi9ra, 
Sraosa, Anahita, and Drvaspa drive chariots (Yast 5.10,30; Yasna 57.27); the 
warrior's weapon includes a sword (Yast 5.130; 10.131; 14.27), a spear (Yast 
10.39,102; 15.48; 17.12), bow and arrows (Yast 7.28; 10.39) and a mace, which 
sometimes has a bull handle, sometimes it has 100 edges (Yast 6.5, 10.96, 101, 
132; 11,10), and sometimes it is made of gold (Yast 10.96, 131). The body was 
covered by protective armor, the main part of which was a helmet made of a bull 
skin or metal (Yast 13.45; 15.57) and a shield. 

Words relating to warriors and the army are Indo-Iranian terms. One of them 
derives from the word 'hero', 'victor', another comes from the verb 'to throw, 
shoot' arrows, spears (Gertsenberg 1972: 33). The name of a missile weapon is 
Indo-European, an arrow name is Indo-Iranian, originating from the verb 'to 
fight', synonyms for the arrow and spear, the axe, adze, shield, protecting 
breastplate are also Indo-Iranian (Gertsenberg 1972: 35, 187, 193, 213, 216, 219, 
221, 222; Pant 1978). 

In the ancient literature the weapon is both a creation and attribute of the 
gods: Indra has the vajra, a stone axe and bull-headed mace with four projec- 
tions; Yima also has a bull-headed mace; BhTmasena possesses a stone mace 
with four projections. The gods present weapons to heroes: Agni gives an arrow 
to Arjuna, Brahma delivers one to Rama, Ahura-Mazdah to Yima. I. M. Steblin- 
Kamensky (1995) suggests that Yima received a sharpened stick. The Iranian 
peoples preserved ideas about the divine origin of weapons for a long time. 
Quintus Curtius Rufus (History of Alexander 7.8.17-18) relates the legend about 
heavenly gifts sent to the Saka, "A plow and bull yoke for farmers, spear and 
arrow for warriors, cup for priests". Herodotus' story about the Scythian gods, 
who guard in fire golden gifts fallen from heaven — a plow, yoke, pole-axe, and 
cup — is a variant of this myth (History 4.5-7). There is a Nart tale about a golden 
weapon falling from heaven. 

The archaeological realia that reflect these mythological ideas about the 
creation of a divine weapon and sky chariot are found only in the culture of the 
Eurasian steppes of the 17th— 16th centuries BC. 

The Aryan arms described in Vedic literature, the Avesta, and in texts from 
Nuzi are similar to ones found in burials of the chariot-warriors along the Don 
and Volga on sites of the Potapovka type and in the Urals, in Sintashta and 
Petrovka burials. In Sintashta were found bronze axes, spears, knives/daggers, 
adzes, hooks, sets of bronze and stone arrows, stone maces, abrasive materials 
for sharpening weapons, single-edged scraper -knives that could be used as a 
currycomb for horses and tools for making wheels such as adzes and chisels, and 
also ornaments, bone handles and cheek-pieces (Gening 1977). In burial 4 of 
kurgan 6 (Utevka IV) a spear, adze, dagger, stone arrows, bone handle-top and 
cheek-pieces were found (Vasil'ev et al. 1992: fig. 4-6). In the Kondrashkinskiy 
kurgan was an axe, spear, knife -dagger, adze, stone arrows, bone cheek-pieces 
and buckles; in Vlas'evskiy kurgan a dagger-knife, stone macehead, bone 
handle-top and buckles were excavated (Vinnikov and Sinyuk 1990: fig. 28; 
Pryakhin 1992: fig. 1). A chariot and a pair of harnessed horses (sometimes 
more) were buried together with warriors in the Urals. This tradition of putting a 
chariot and horses in a chiefs grave also continued in the steppes in the Scythian 



138 CHAPTER EIGHT 

period, which is demonstrated by the discovery of vehicle parts and handles in 
the Elisavetinskiy kurgans, Melitopol', Tolstaya Mogila, Gaimanova Mogila, 
Chertomlyk, Aleksandropol', Krasnokutsk and Vasurinskiy kurgans. The axle 
width of Scythian chariots is similar to that of the Andronovo; the wheel 
diameter is smaller, the spokes number 8-12 (Rostovtsev 1914: table xii.l; 
Terenozhkin 1972: 123; Mozolevsky 1979: 191ff). Vehicles and horses were 
found in the eastern part of the steppes, in Pazyryk (Rudenko 1953: 146-235), 
moreover, in all cases they were found only in royal kurgans. 

These data confirm the identification of the Andronovo tribes with the Indo- 
Iranians. Arguments about the formation of Central Asian metal cheek-pieces of 
Class I and II under the influence of more archaic bone cheek-pieces of the Eur- 
asian steppes also supports this identification. The evidence pertaining to the his- 
tory of the horse in the Old World also proves it. Paleozoologists V. I. Gromova 
(1940; 1949), V. I. Tsalkin (1970) and V. I. Bibikova (1967; 1970) maintained 
that horse-breeding became one of the main branches of the steppe economy 
already from the middle of the 4th millennium BC. Religious and mythological 
beliefs associated with the horse emerged in this zone in the same period. In 
most ancient ritual horse burials of the Old World on sites of the Mariupol' and 
Pit-grave cultures one can find sources of the asvamedha and other horse cult 
rituals characteristic of the early Indo-Europeans and Indo-Iranians in particular 
(Ivanov 1974; Kuz'mina 1977; Mallory 1981). Paleozoologists attribute the Near 
Eastern chariot horses to the same breed as Andronovo horses: they were 
brought from the north, from the steppes (Azzaroli 1975). 

Thus, the evidence of Andronovo transport history provides an important 
argument in favor of the Indo-Iranian attribution of the Andronovans. 



In the 12th century BC chariot warfare tactics lost their importance in Androno- 
vo society; mounted horsemen armed with bows and arrows replaced chariot 
drivers. The bottom date for the change in battle tactics over the broad steppe 
zone is set to the 12th century BC because it is then that we see the spread of 
cheek-piece types which genetically preceded those of the Scythian horsemen. 
Mounted battle tactics required a steady seat for a horseman. It must not be 
confused with riding as it was practiced by pastoralists from the moment of horse 
domestication (Kovalevskaya 1976). In the Near East individual images of 
unarmed riders in the last quarter of the 2nd/beginning 1 st millennium BC are 
known in Amarna, Sakkar, Sindzherli, on a seal from the Louvre (Schafer and 
Andrae 1925: 381; Yadin 1963: fig. 219-224; Meyer 1953: 32; Herzfeld 1937: 
51; Azzaroli 1985:23). 

At the same time horse-riding spread to Greece, Cyprus, Anatolia, and the 
Caucasus which is proved by numerous images of riders in applied art (Anderson 
1961; Bosert 1942: 91, fig. 945, 1183; Hanfman 1961; Uvarova 1900: tables 
lx,2, lxi,8; Hancar 1935: table 14; Kuz'mina 1973c). 

Later mounted warfare appeared in the Near East. Written texts show that 
especially in the first centuries of the 1st millennium BC riding became more and 
more important in the armies of Assyria and Israel (Kings 1.5, 4.26, 10.26) (VDI 
1951: no. 2, 296, 299; no. 3, 330; Yadin 1963: 184-186; Saggs 1963: 2). This 



TRANSPORT 139 

process is reflected in Assyrian art: on monuments of the 1st millennium BC the 
images of riders become more and more numerous (Herzfeld 1938: 50-51; 
Barnett 1960; Barnett, Falkner 1962, Hanfman 1961: 243-255; Yadin 1963: 415- 
417; 442-443, 450-451). It is seen from written documents that the appearance of 
riders in the army was necessitated by defensive measures against the raids of 
steppe tribes from the north. Scholars agree that this innovation in warfare was 
influenced by the steppe peoples. 

M. Gimbutas' hypothesis (1956) about the emergence of horse riding in the 
Eurasian steppes already in the Eneolithic and the role of these horsemen who 
destroyed European cultures with sword and fire has been supported by V. N. 
Danilenko and N. N. Shmagly (1972) and D. Anthony (1995) finds no proof. In 
the first place, as N. Ya. Merpert (1961b: 77-80; 1966b) and E. Comsa (1976) 
convincingly showed, there was no mass invasion of steppe peoples into Europe, 
rather there was gradual penetration by separate groups of population from the 
south Russian steppes; moreover, none of them was an all shattering invasion. In 
the second place, objects that have been described as cheek-pieces were not 
really ones (the same is true of Siberian objects understood as cheek-pieces by P. 
M. Kozhin). The function of the latter is defined on the analogy with Anyang 
objects as devices for knot unfastening (Komissarov 1980) or net braiding which 
is proved by findings of analogous objects in European cultures that did not 
know the horse (Dietz 1992). 

The late appearance of horse riding is affirmed by Indo-European linguistic 
data as well. In early Indian literature, in the Rigveda and Mahabharata, the 
chariot is ubiquitous but horse riding is mentioned only a few times (Kal'yanov 
1967: 136-160; Kane 1946: 200-205; Pusalker 1963). In the ancient parts of the 
Avesta the common Indo-Iranian name of the social group of raOaestar 'chariot 
driver' is used. Only in Yasna 11.2, a relatively late text, is the term 'chariot 
driver' replaced by the word for 'horseman', basar (Benveniste 1951: 122-124) 
which has no Indian correspondences. As this text was compiled certainly not 
before the beginning of the 1st millennium BC it proves that originally the Indo- 
Iranians had chariot-warfare tactics and horsemanship appeared only at the end 
of the 2nd millennium BC. 

Similarly in Mycenaean Greece a member of military aristocracy was called 
hippeus, 'chariot driver'. The term 'charioteer' was preserved even when battle 
tactics had changed and Homeric heroes drove chariots onto the battle-fields to 
fight on foot; in the 6th century BC this term was used to denote a privileged 
social group of horsemen (Lurie 1957: 324-330). 

Names for wheeled transport and the horse are common in all Indo-European 
languages; they all have similar myths and ideas about sky and solar gods 
driving chariots across the heavens; Indo-European driving and wheel burning 
rites on solar holidays are similar, but names connected with riding do not reflect 
Indo-European and Indo-Iranian unity. 

V. I. Abaev (1979 III: 357) showed that the Ossetic words connected with 
vehicle and chariot have a common Indo-European and Indo-Iranian origin in 
particular, while the whole terminology connected with riding is of Iranian origin 
(rider, bridle, horse cloth, saddle-girth, tail belt, etc). 

The linguistic evidence confirms that the Indo-Europeans, as elsewhere in the 
Old World, originally practiced chariot -warfare and that horse riding was spread 



140 CHAPTER EIGHT 

later. In the period of the spread of the horse-drawn chariot the Indo-Europeans 
still had contacts with each other; moreover, active contacts can be traced be- 
tween Greeks and Indo-Iranians who share a number of common mythologems. 
Horse riding was completed only after the collapse of Indo-Iranian unity. 
According to archaeological data Indo-Iranian unity should finally dissolve not 
before the 17th— 16th century BC, but earlier than the 12th century BC (or if we 
take into account calibrated data - not before beginning of the 2nd millennium 
BC). 

Evidence for the emergence of horse riding in the 12th century BC is 
important for resolving the problem of the migration of the Iranian-speaking 
peoples. E. A. Grantovsky (1970) maintained that analysis of Iranian names 
indicated that they appeared on the Iranian plateau at the end of the 2nd millen- 
nium BC and became numerous only in the 8th century BC. In Iranian culture of 
this period there appear a number of important innovations. At Hasanlu, Dinka- 
Tepe, Marlik, and Babadzhan archaeologists have discovered a horse burial rite 
that is alien to early Iranian cultures (Dyson 1965: 210-211; 1967; Ghirshman 
1964: 56; 1963: 24, 26, 28; Muscarella 1968: 35; Negahban 1964: 15, 16) and in 
Luristan and Sialk VI we have a horse-harness burial rite (Maleki 1964: 35; 
Vanden Berghe 1959: 104-123; 1968; Ghirshman 1963: 279). Numerous images 
of riders, horses, and griffons found in Amlash are marked in art (7000 ans: 25, 
no. 130, 131; Calmeyer 1964: table 10, 26; Gabus 1967: table xvii), Luristan 
(Hancar 1934: fig. 38; 1935; Potratz 1942: fig. 12-14, 40), Sialk VI (Ghirshman 
1939: 1; 1963: fig. 69, 75). In Sialk VI and Giyan a bronze bit of pre-Scythian 
type (Class III, type II) was found. R. Ghirshman (1939) was the first to 
associate these innovations that he found in Sialk with the migration of the 
Iranians. In his last work R. Ghirshman (1977) unfortunately rejected this 
hypothesis (for a critical approach to his new ideas see Lelekov 1978: 220-226). 
The participation of the Medes and Scythians in the creation of the Luristan 
culture (Vanden Berghe 1968) or that of the Cimmerians (Ghirshman 1963: 42, 
71-72) as well as R. Dyson's (1965: 208-209) attempt to connect the horse 
burials in Hasanlu with the Cimmerians and Scythians are not convincing due to 
the later arrival of the latter (Kuz'mina 1973c: 185, 187; 1977: 33-34; 1971d; 
Pogrebova 1971: 258-260; 1977a: 114-140; 1977b). But their connection with 
the migration of the Iranians is probable. This is affirmed by the fact that these 
new features in the spiritual culture of Iran have a long tradition of development 
in the Eurasian steppes (Kuz'mina 1971d; 1973c; 1977; Pogrebova 1977a; 
Mallory 1981). These ritual burials and images of horse and riders probably 
mark the route of Iranian-speaking tribes moving from the Eurasian steppes to 
the Iranian plateau who brought with them horse raising, riding, mounted combat 
and the cult of the horse. 



CHAPTER NINE 
ECONOMY 

The economy of the Andronovo culture was based on mixed farming and 
stockbreeding with different emphasis on the various components according to 
both chronology and region. 

Farming 

O. A. Krivtsova-Grakova (1948: 103-104) and S. V. Kiselev (1949: 56) were the 
first to study Andronovo farming. The main indirect evidence of farming is the 
topography of the Andronovo settlements, the majority of which are situated on 
fertile grounds in the flood-plains of rivers (Sal'nikov 1951b: 126; 1954: 248). 
There is little direct evidence for Andronovo farming. At Alekseevka on a ritual 
mound there were pits with carbonized stalks and grains of wheat (Krivtsova- 
Grakova 1948: 73). There are old identifications of rye (?) and millet found on 
settlements in Kazakhstan (Minaeva and Furaev 1934; IK SSR 1977: 111) and 
palaeo-botanical discoveries of wheat and millet on settlements of the Fedorovo 
type and the Cherkaskul' culture (Lebedeva 1996: 54, 55) and neighboring Tim- 
ber-grave sites of the Urals (Lavrushin and Spiridonova 1999: 102). There is the 
imprint of cereals preserved on the bronze sickle from a settlement at Malokras- 
noyarka (Chernikov 1960: 44, table 36, 19). Traces of burnt roofing material 
were found at the settlement of Ushkatta 2, consisting of clay-covered straw (the 
cereal involved was indeterminate to species; excavations by E. E. Kuz'mina). 

Primitive stone querns are found on the majority of Andronovo settlements; 
they are rectangular or oval, 15-25cm long, 2-6cm thick, with artificially deep- 
ened surfaces. There are three types of grinding stones: elongated, rectangular in 
section, round or cylindrical, sometimes with a conical handle, or shaped like a 
pyramid with a flattened top: Alekseevka, Kipel', Novo-Burino, Zamaraevo, 
Kambulat 2, Ushkatta 2, Kiimbay, Elenovka, Shandasha, Atasu, Buguly 1, 
Suukbulak Milykuduk Tagibay-Bulak, Shortandy-Bulak, Karkaralinsk, Ust'- 
Narym, Malokrasnoyarka, Trushnikovo, settlements of northern Kazakhstan 
(Krivtsova-Grakova 1948: 103-104, fig. 18, 30; Sal'nikov 1951b: 125, fig. 12; 
1967: 333; Kuz'mina 1962a: 88; 1962b: 14, fig. 3, 6; 1964b: 106; Margulan 
1979: 262, table 8.1, fig. 126, 134, 162, 164, 165, 172, 180; Chernikov 1960: 37, 
tables 25, 26.2, 27.13, 39.3,4,7, 50,5, 51; 52; Zdanovich 1979: 11, 18; Chebako- 
va 1975: fig. 3.15, 16; 4. 12; 5. 8). There are interesting finds of querns in the 
ritual complex of Baksay on the Atasu river in central Kazakhstan and in burials 
in the Urals at Adamovka and Ataken-say, in the latter case under the head of a 
girl (Margulan et al. 1966: 260; Sal'nikov 1967: 333; Kuz'mina 1963b: 127). 
The finding of cereal grains in a ritual context at Alekseevka and the sickle in a 
woman's grave at Tulaykin Aul attest to a farming cult among the Andronovans. 
Judging by the sickle and querns in women's burials the harvesting and 
processing of grains was a female occupation among the Andronovans. 



142 CHAPTER NINE 

However, querns and pestles need not have been used only for grain grinding but 
also for crushing wild plants and ore (such use is supported by assemblages in 
industrial complexes at Shandasha, Atasu, Dzhazkazgan, Trushnikovo and near 
mining sites, cf Kuz'mina 1964b: 106; Margulan 1979: fig. 126, 180; Chernikov 
1960: table 2; 1949: tables 1.1,2,3, 5. 1), which has been supported by use-wear 
analysis by G. F. Korobkova. 

Finds of bronze sickles of several types are known over the entire Andronovo 
region. The first group comprises sickle-like knives with a slightly concave blade 
without a defined handle. These appear already in the Sintashta-Petrovka stage 
(the settlements of Arkaim and Kulevchi, the Sintashta burial), they are known in 
the Alakul' and Kozhumberdy complexes (Tulaykin Aul burial, the settlements 
of Chernyaki 3, Kambulat 2, Starikovskoe, Ushkatta 8, Kamyshnoe 1, Petrovka 
2, Novonikol'skoe, Bogolyubovo 1) and survive to the Final Bronze Age 
(Stalinskiy Rudnik, Myrzhik) when they become numerous on sites of the steppe 
cultures of Central Asia and also appear in western Siberia (Grakov 1935: fig. 
67.6; Gryaznov 1956: 30-31, fig. 7, 26; Orazbaev 1958: 276, 278, table 10, 2; 
Chernikov 1960: table 8.4, 5; Kuz'mina 1966: 44-46, table 9; Sal'nikov 1967: 
331, fig. 53.7, 8; Zdanovich 1973: 118; Chebakova 1975: fig. 5.10.11; Stokolos 
1972: fig. 6. 1; Margulan 1979: 5-9, fig. 166; Avanesova 1979: 14; Zdanovich 
1988: tables 10, 20, 21; Kadyrbaev et al. 1992: fig. 118. 11; Gening et al. 1992; 
fig. 75.6; 148, 164; Zdanovich 1997: fig. 9). Early sickle-shaped knives were 
polyfunctional and could be used as knives, sickles, and carpenter planes. These 
tools are quite analogous to Abashevo and Timber-grave ones and are genetically 
connected with them (Krivtsova-Grakova 1955: 54, fig. 12.12; Sal'nikov 1967: 
fig. 2, 3, 5, 6.1-5; 24.12; Chernykh 1970: fig. 54. 12-17, 55. 1-9). Their forms 
developed from slightly curved knife -sickles to a specialized type of knife with a 
straight blade and then a sickle itself. Its efficiency rose due to an increase in the 
curvature of the blade. At the end of the Bronze Age sickles often had holes for 
attaching a handle. Andronovo mass production of sickle-shaped knives is shown 
by matrices with some negatives in the Urals at Kundravinskaya and at Orenburg 
(Krivtsova-Grakova 1948: 103; Tikhonov 1960: table 20. 14). 

The second group consists of sickle-choppers of the Sosnova-Maza type (Fig. 
43: 19-21) found in the Urals, north, central, and eastern Kazakhstan, Semi- 
rech'e and in west Siberia, including those found on settlements of the Final 
Bronze Age at Stepnyak, Alekseevka, Yazevo, Ust'-Narym, Malokrasnoyarka, 
Konezavod, Chaglinka, in the burial at Zevakino, in the late layer of Petrovka 2 
and Novonikol'skoe, and in the hoards of Alekseevskiy, Turksib and Shamshi at 
the end of the 2nd/beginning of the 1st millennium BC (Chernikov 1949: table 
10.4; 1960: 38, 44, 162, tables 36.19, 16.11; Orazbaev 1958: 141; Akishev and 
Kushaev 1963: 108, fig. 83; Grishin 1960: 123, fig. 1, 6-7, 20; Chlenova 1955: 
fig. 3, 5; AO 1969: 393; 1973: 467; Kuz'mina 1966: 54-56, table 11; Kozhum- 
berdiev and Kuz'mina 1980: 145-146, fig. 1, 17-19; Krivtsova-Grakova 1948: 
107, fig. 27.3; Evdokimov 1975c: 112; IK SSR: fig. 108.1, 2; Arslanova 1974a: 
tables 1, 4; Zdanovich 1973a: 118; Avanesova 1979: 15; Potemkina 1979: fig. 
6.3). The mass production and late date of sickle -choppers is affirmed by a 
bronze mold (Omsk museum). V. V. Golmsten thought that these tools were not 
sickles but cutting instruments used for clearing flood land for sowing and for 
stocking up branches for cattle. Modern Turkmen use an analogous tool for 



ECONOMY 143 

slaughtering cattle, so it is possible that sickle -choppers were connected rather 
with cattle-breeding and not so much with farming. Andronovo sickle-choppers 
are analogous to those of the Timber-grave culture (Sosnova-Maza, settlement at 
Samara, Osinovye Yamy) and share a common genesis with them (IAK 1909: 
vol. 29: 65-66, fig. 6, 11; Krivtsova-Grakova 1955: 62, fig. 14.7; Merpert 1966a: 
132; Chernykh 1970: fig. 55, 10-15). 

The third group is composed of sickles with a hook known in the Urals, 
northern and eastern Kazakhstan including the Final Bronze Age settlement of 
Yavlenka (Zdanovich 1988: table 10c, 110), and is analogous to a similar form 
from the Timber -grave culture. 

The fourth group consists of sickles and scythes with one or two flanges; they 
are known in northern and eastern Kazakhstan, Semirech'e, in western Siberia 
including the Final Bronze Age hoards of Issyk-Kul' and Predgornoe, in a burial 
at Predgornoe and on the settlement of Chaglinka together with pots with 
applied-roller decoration that indicate the date of this type (Kuz'mina 1961c; 
1966: 56-57, table 11. 9, 10, 15-17; 1967: fig. 1; Kozhomberdiev and Kuz'mina 
1980: 86-87, fig. 1, 10; Grishin 1960: 123, fig. 1, 8; IK SSR: fig. 108, 4; Oraz- 
baev 1970: 141). We can see then that harvesting tools appeared in the Androno- 
vo culture from its formation and shared a common genesis with tools from 
Eastern Europe, Abashevo and Timber-grave, and developed along similar lines 
over the whole Eurasian steppe. In the Final Bronze Age the number of sickles 
increased and local improvements were developed, which points to the 
development of farming in various isolated groups. 

The nature of earth moving tools is not clear. Archaeologists believe that the 
Andronovans had hoe farming (Krivtsova-Grakova 1948: 103; Sal'nikov 1951: 
124; Chemikov 1960; Margulan et al. 1966: 260-261; Markov 1973: 104; IK 
SSR: 111; Potemkina 1985: 318). Bronze adzes could have been used as earth- 
moving tools. They were known in the west of the Andronovo territory: the 
burials of Novy Kumak, Kenes, Petrovka (Smirnov and Kuz'mina 1977: 36, fig. 
3.2; Zdanovich G and S. 1980: 183, fig. 1, 7), in burials of the Alakul' complex 
at Orsk, on the Emba and on settlements at Chelkar (Formozov 1951a: 120, fig. 
32; Kuz'mina 1961b: 92-93, fig. 32.10; Chernikov 1951: 159, fig. 2, 10; 
Orazbaev 1958: 275, table 11, 9-11, Gening et al. 1992, fig. 127: 140, 146, 148, 
152, 164, 175, 184, table 39). In eastern Kazakhstan, the Minusinsk region and 
the Central Asian steppes only individual cases of such findings are known 
(Chernikov 1949: 67, table 7.2; Martin 1883, table 8, 9; Kuz'mina 1966: 17, 
table 3, 12). Andronovo adzes are analogous to Abashevo and Timber-grave 
adzes and they have a common genesis (Krivtsova-Grakova 1955: fig. 13, 3-5; 
Chernykh 1970: fig. 52.4-9, 11-14; Sal'nikov 1967: fig. 6, 15, 24.4; Gorbunov 
1986: table 15.9, 10). 

On the Tobol, and northern, central and eastern Kazakhstan, in Semirech'e, 
in the Altai and western Siberia there were adzes of another type — massive 
palstaves with flanges (Fig. 43a: 11; 43b: 6-9). They have been recovered on the 
settlements and in houses of the Final Bronze Age at Petrovka, 2, Alekseevka (a 
mold) and in Malokrasnoyarka; they are also included in numerous hoards dating 
to the end of the 2nd millennium BC. These are Sary-Ozek, Kamennoe plateau, 
Alekseevskiy, Sukuluk, Sadovoe, Karakol 2, Shamshi and Balandino (Tallgren 
1911: 123, fig. 71; Gryaznov 1927: 211: 1930: 162, 156, fig. 3, 13, 14; 



144 CHAPTER NINE 

Chernikov 1949: table 8.2; 1960: 44, 82, 164; table 36.7, 64.9, 67.5 
Moshinskaya 1957: 61.5; Kuz'mina 1966: 18-20, table 3, 9, 10, 14-17; 1968a 
13-15; Kozhumberdiev and Kuz'mina 1980: 144, fig. 1, 25; Evdokimov 1975c 
112; Zdanovich S. 1974b: 320; 1979: 12; Margulan 1979: fig. 2, 4; Zdanovich 
1988: table lOd, 18). Andronovans used adzes of both types: traces of tool use 
have been seen on both the foundation ditches of their houses and grave pits. 
However, for the cultivation of large fields, adzes, which have very narrow 
blades, were of little use. Woodworking was the main purpose of this multi- 
functional tool. Celts of the Seyma type, most probably of Andronovo origin, 
were used for digging (Gryaznov 1941: 251-254; Zotova 1964: 59-63; 
Chernikov 1960: 80, 81), but there are no such finds in complexes at present. 
Celt-spades appeared at the end of the Bronze Age in central and eastern 
Kazakhstan, in Semirech'e, Fergana, and western Siberia (Chernikov 1951: table 

1, 17; 1960: 83, table 15. 4; Grishin 1960: 122, table 7.4; Kuz'mina 1961b: 258, 
259; 1966: 24-26, table 5; Margulan 1979, fig. 2, 3). 

Stone and horn tools could also have been used for loosening the soil 
(Zdanovich, S. 1979: 18, Kadyrbaev et al. 1992, fig. 116) but their use would 
have been limited to secondary cultivation of the soil. 

Stone hoes are considered to be the main agricultural tool of the 
Andronovans. They are found across the whole Andronovo region on settlements 
such as Alekseevka, Zamaraevo, Elenovka, Kiimbay 1-7, Ushkatta 2, Shandasha 

2, Atasu, Myrzhik, Ak-Mustafa, Ulutau, Karkaralinsk, Buguly, Suukbulak 
Tagibay-Bulak, Milykuduk Kanay, Ust'-Narym, Trushnikovo, Semipalatinskie 
Dyuny, and on settlements of northern Kazakhstan. They are dated to the Full 
and Final Bronze Age (Krivtsova-Grakova 1948: 103, fig. 28.1, 2; Sal'nikov 
1951b: 124-125; 1967: 331; Sal'nikov and Novichenko 1962: fig. 6; Kuz'mina 
1962a: 14, fig. 3, 8; 1964b: 106; Margulan et al. 1966: 260, tables 23, 37.1 
33.15, 43.1, 47; Margulan 1979: 171, 215, 231, fig. 126, 164.5, 173, 180, , 4, 5 
Chernikov 1960: 29, 37, 56, 88, tables 16, 26.4, 5, 49; Zdanovich, S. 1979: 11 
Potemkina 1985: 318.1; Kadyrbaev et al. 1992: 149-155, fig. 113-115). These 
tools are nearly rectangular or trapezoidal in form with slightly curved blades 
and hollowed out for fastening onto an elbow-shaped wooden handle. They are 
made of solid rock and often well polished. According to S. A. Semenov, the 
Ust'-Narym hoes were used for breaking up the soil (Chernikov 1960: 556). 
However, use-wear analysis of seven samples from the Elenovka micro-district, 
undertaken by G. F. Korobkova, showed that they were used as mining 
instruments, which is proved by the split character and worn out traces formed 
by hard rock strokes and the presence of copper ore fragments. The distribution 
of these 'hoes' near ancient mining districts and their discovery in the Temir and 
Sarybulak mines (Kadurbaev et al. 1992: 150) support this conclusion. 

All the above-mentioned tools, even if we accept that they were used for 
cultivation, could be used only for small-scale gardening of the flood lands; for 
field cultivation they could not be used as the soils characteristic of the Urals and 
Kazakhstan demand deep plowing. Consequently, if Andronovo agriculture was 
hoe farming, it had a low efficiency and could not be significant for the 
economy. However, the abundance of grain storage pits within the settlements 
and the general level of cultural development among the Andronovans, who 
widely employed domestic animals, suggests that they also used the plow. But 



ECONOMY 145 

on the petroglyphs of Kazakhstan, in contrast to those of Central Asia, the plow 
is not evident, although figures of vehicles and chariots are often met. Thus, the 
question of the tool usage and ways of land cultivation of the Andronovans and 
thus of their farming character and its scale is not quite clear although arable 
farming does seem highly probable. 

The distribution of Andronovo settlements suggests some regional economic 
variation (Maps 13, 14). If in the north of their territory, in the forest-steppe zone 
of the Urals and northern Kazakhstan, settlements are situated along the river 
banks near flood plains on fertile lands suitable for hoe-gardening and arable 
field farming, in the southern steppe zone in the south Urals, western and 
especially central Kazakhstan many sites are on the open steppe with stony and 
saline soils and even in semi-desert regions where cereal cultivation was 
inefficient or impossible (Sorokin 1962a: 59; Kuz'mina 1962b: 92; Margulan 
1979: 261). This conclusion is supported by the fact that in settlements of the 
forest-steppe in every house there are several, sometimes dozens, of pits for 
grain, in which grain could be kept during the whole winter, without losing its 
germinating capacity (Bow en and Wood 1968), while in the southern Urals and 
western and central Kazakhstan they are found very rarely or not at all, and the 
only pits discovered are associated with storing ore. In these fundamental 
districts of the Andronovo culture only stock-breeding could form the basis of 
the economy, making the Andronovo economy mainly a stock-breeding one. 

There are indications that at the end of the Bronze Age the Andronovans 
constructed artificial dams and employed irrigation from streams and estuaries, 
resembling Tazabagyab ones (Andrianov 1969; Itina 1977: 176-178), but it was 
more primitive. Dams and small water courses have been discovered in the 
southern Urals at Lake Buruktal, at Elenovka and in northern Kazakhstan 
(Sal'nikov 1967: 333-334; Zdanovich, S. 1979: 19). In central Kazakhstan large 
dams are known, made of two rows of stone slabs and an earthen embankment 
partitioning off gorges and broad gullies of the Altyn-Kazgan, Keregetas, 
Kipchakpay, Korgantas, Kushoku, Milykuduk, Akchi, etc. (Margulan 1979: 263- 
270, fig. 188-197). Irrigation could have been made for field irrigation and the 
creation of water-meadows, but dams in the Elenovka micro-district and 
numerous ones from central Kazakhstan suggest by their topography and traces 
of flotation minerals that they were used not for amelioration of the fields but 
rather for washing ore. 

Andronovo irrigation works are fundamentally different in structure from the 
contemporary and more ancient irrigation systems of south Central Asia and 
Iran, which proves their local origin. All types of Andronovo agricultural tools 
also have nothing in common with those of Central Asia and Iran, but they are 
identical to those of Eastern Europe. 

This indicates a similar level of development and a common genesis for the 
farming economy of the Andronovans and Timber-grave people in Kazakhstan 
and western Siberia, because proto-types of many of the common steppe farming 
tools are known in Eastern European complexes of the pre-Seyma period 
(Poltavka and Abashevo). 



146 CHAPTER NINE 



Stockbreeding 



The formation of the economic-cultural type and the evolution of the economy of 
the steppe tribes were to a large extent defined by the natural and climatic 
conditions of the steppes and their periodic changes. 

The Eurasian steppes belong to the zone of unstable farming. Although 
chernozems (productive black earths) constitute more than 2.5 million hectares, 
large areas comprise soils of low fertility, grey soils, white saline soils and sand 
(Milkov 1964). The climate of the steppe is severely continental. The amount of 
precipitation from the northern borders of the steppe to the southern falls from 
430mm to 150mm per year, and the moisture coefficient is six times greater in 
the south due to evaporation. Some districts in the south and east become a semi- 
desert zone (Mordkovich 1982: 26). Although Mesopotamian field irrigation 
produced crops that yielded thirty fold and more, in Turkmenia fifteen fold, in 
Greece 800-1000kg per hectare, in the Balkans 225-450kg per hectare (Masson 
1971: 102), in western Kazakhstan even in the 19th century the yield of the 
average crop was only four fold. But for the development of stock-breeding the 
steppes offered optimum conditions, providing natural pasture. In the steppe 
characterized by feather-grass and various other grasses it was possible to obtain 
up to 15 quintals of hay per hectare, in wormwood-tipchak steppe up to 7 
quintals, and in semi-desert up to 5 (Milkov 1964: 167-169). Six-seven heads of 
oxen and horses could be pastured on 1 square km of mixed-grass steppe (Mord- 
kovich 1982: 185). This preconditioned the further evolution of the economy by 
increasing the role of stock-breeding. 

A complex sedentary economy was established in the steppes in the 3rd 
millennium BC (Kuz'mina 1996). The next stage in the evolution of the 
economy was in the 17th 16th centuries BC, the period of the Timber-grave and 
Andronovo cultures. The economy remained complex sedentary which is 
indicated by fortified settlements of the Arkaim type. However, Petrovka tribes 
did not choose the path toward urban civilization, the intensification of the 
economy and growth in labor productivity. The specific character of its 
ecological niche conditioned the development of Andronovo society. 

The stock-breeding of the Sintashta population derived from Abashevo. They 
raised very tall cattle, sheep and mainly horses which were exploited for food 
and for the first time for pulling chariots; the role of the pig decreased sharply 
(Varov and Kosintsev 1996: 54). Bones of domestic animals constitute 96% of 
the faunal remains of which cattle is 60.4%, ovicaprids 24.2%, horse 15.4%, and 
single bones of pig or boar are encountered (Zdanovich 1997: 56, 57). The same 
composition of species is recorded for Sintashta-Petrovka burials. Hunting 
played a minor role, but the combination of steppe and forest species is of 
interest: wolf, fox, hare, marmot, boar, roe deer, saiga, aurochs, bear, elk red 
deer (Zdanovich 1975: 57). Two-humped camels {camelus Bactrianus), probably 
domesticated, appeared for the first time (Lavrushin et al. 1999: 101). These 
tendencies developed in the next stage, marked by the apogee of the culture seen 
in the expansion of territory and the abundance of sites which surpassed the 
number of other periods. 

This was the time of the climate optimum of the Subboreal period when the 
climate became milder. Carbon analysis of the Elenovka micro-district, 
conducted by G. N. Lisitsyna, and geological and paleoecological research of the 



ECONOMY 147 

Arkaim area (Lavrushin et al. 1999: 101-102) demonstrate that there appeared 
forest regions in the steppe, where birch, lime, aspen and pine grew, and under 
the influence of Siberia the Siberian pine, larch, and fir-tree appeared, which do 
not grow here at the present. Andronovo houses and the timber-frameworks in 
graves were made of pine. Various grasses and cereals dominated in the steppe. 
These conclusions are confirmed by the composition of wild animals, which 
include both representatives of steppe fauna: wolf, fox, hare, boar, kulan, saiga, 
roe -deer, argali-djeyran, and forest animals: steppe bear, stag and reindeer, elk 
and beaver, that are found across the whole Andronovo natural habitat, although 
not met here at the present (Smirnov 1975; Khabdulina et al. 1984: Nurumov 
1987; Akhinzhanov 1992). Paleoecological conditions were favorable for the 
spread of a complex economy with stock-breeding dominant. It provided the 
Andronovans with meat, milk, wool, skins, fuel and bone for implements. 

The composition of the herd is reconstructed on the basis of faunal remains 
from settlements. Cattle occupied first place, ovicaprids second, and horses were 
third. Dog and the Bactrian camel were also known (Krivtsova-Grakova 1948 
100-102; Kiselev 1949: 55-56; Sal'nikov 1951b: 122-124; 1954a: 148; 1967 
327-330; Gryaznov 1957: 21-28; Chernikov 1960: 62, 64, 167; Kuz'mina 1963a: 
38-46; 1977: 28-52; 1980: 11-35; Margulan et al. 1979: 257-263; Makarova 
1974: 201-206; 1980: 141-151; IK SSR 107-109; Zdanovich S. 1979: 17-18; 
1981: 44-54; Maksimenkov 1978: 48, 51; Akhinzhanov et al. 1992). 

The specific character of Andronovo stockbreeding was examined by V. I. 
Tsalkin (1964: 24-30; 19972b: 66-81) and furthered by P. A. Kosintsev (1989), 
Varov and Kosintsev (1996) and S. M. Akhinzhanov (1992). Andronovans raised 
the same breed of cattle (Ukrainian grey type) that was found in the Abashevo, 
Timber-grave and earlier Eneolithic cultures of the south Russian steppes and 
along the Danube. It were large (withers height was 126cm, average weight was 
350kg) and long horned animals but many were hornless (Tsalkin 1972b: 70-71). 
In the Trans-Urals and eastern Kazakhstan a smaller breed of cattle was found 
(Krivtsova-Grakova 1948: 102; Sal'nikov 1951b: 128; 1967: 328; Chernikov 
1960: 62; Kosintsev 1989: 94). In addition to oxen large bulls were used for draft 
(withers height 133cm; Tsalkin, 1964: 26; 1972b: 71, 72). The use of a bull 
harness by the Andronovans is confirmed by Kazakhstan petroglyphs. 
Andronovans sacrificed mainly bulls rather than cows; usually large mature 
animal skins with a skull and legs, substituting for the whole animal under the 
principle of pars pro toto, were placed in burials (Tsalkin 1972b: 72). 
Andronovo stockbreeding was mainly meat oriented which is indicated by 
evidence for butchery on the bones among the food waste. In the western 
Andronovo region young animals were mainly used for food; 50% of slaughtered 
animals were not older than two years, 75% were younger than three years. This 
yielded a highly productive strategy of livestock slaughter. 

The Andronovans also consumed milk from cows. The main evidence are 
vessels in all burials and cult places. Analysis of their content has not been 
conducted, but they probably contained vegetable and dairy food (Gryaznov 
1957: 23). The meat and dairy focus of the economy, especially in the forest- 
steppe, is indicated by the age-slaughter pattern (Kosintsev 1989: 96, table 5). In 
the Final Bronze Age they undoubtedly used dairy products and learned to 
process them: on the settlement of Kipel' a vessel for cheese and cottage cheese 



148 CHAPTER NINE 

making, with holes in its base, was found (Sal'nikov 1951b: 124). Cheese 
production supplies products for long term storage and was an important 
invention necessary for a nomadic and mountain pasture economy, where cattle 
were far from the settlement for a long time. 

The Andronovans processed skins of all domestic animals, using blunt knives 
fashioned from jaw bones. These are numerous on settlement sites (Krivtsova- 
Grakova 1948: fig. 24; Margulan et al. 1966: table 3; Chernikov 1960: 28-29, 46, 
tables 15.1; 38.2; 39.5). 

Ovicaprids occupied second place in the Andronovan herd. The Andronovans 
raised the same breed of sheep as the Timber-grave and other synchronous and 
earlier cultures of the south Russian steppes. It was a large sheep with an average 
height at the withers of 70cm and more; its weight was 50kg (Tsalkin 1972b: 72- 
73; Galchenko 1990: 60; Kosintsev 1989: 98; Akhinzhanov et al. 1992). Judging 
by remains from burials, sheep were sacrificial animals. Sheep were used for 
meat (there are many bone fragments on settlements). Young animals were 
mainly used for food: two-thirds of the bones belong to individuals up to one 
year and a half (Tsalkin 1972 b: 72). According to A. V. Galchenko (1990: 60) 
most sheep remains derived from animals from five months to a year in age. He 
connects this with the fact that lamb skins of this age are the most suitable for 
making sheepskins. Sheep wool was spun: many clay spinners, twisted woolen 
threads (Alekseevka, etc.) and knit objects (Orak, Andronovo, Ust'-Erba, 
Pristan') as well as cloth and impressions on ceramics are known (Sosnovsky 
1934: 92-96; Krivtsova-Grakova 1948: 101; Kiselev 1949: 42, 48; Maksimenkov 
1978: 14, 40, 72). According to statistical analysis of osteological materials from 
settlements ovicaprids occupied second place in terms of the number of bones 
and individuals, but often they come first. However, even though sheep are more 
prolific (3-5 lambs per year) than cows (1 calf) and may account for a greater 
proportion of the bones or individuals recovered (Gryaznov 1957: 23), since 
sheep weigh one-seventh that of a cow it turns out that in spite of the large 
number of sheep bones on the settlements the food of the Andronovans consisted 
of 60-70% beef and only 10% of mutton (Tsalkin 1972b: 80). This indicates a 
real but not a very significant role for sheep-breeding in the Andronovo 
economy. By following horses sheep could obtain food in the winter under snow 
(Gryaznov 1957: 25; IK SSR: 107) which was a main precondition for the 
transition to mountain pasture and semi-nomadic stockbreeding. 

The Andronovans raised goats, but these played an insignificant role. Goat 
bones have never been found in burials, consequently they did not belong to the 
sacrificial animals (there are exceptions in the burials at Hripunovo and 
Chistolebyazh'e). Andronovo goats as well as the goats of the Timber-grave and 
other East European cultures belong to the prisca-aegagrus type. Falconeriod 
goats have not been recovered (Tsalkin 1972b: 73). Thus, both types of 
ovicaprids and the insignificant role of the goat in the economy indicate a 
principal difference between the Andronovo stockbreeding and that of south 
Central Asia and Iran, where sheep-breeding and goat-breeding played the main 
role, thus providing another important proof of the East European and not south 
Central Asian origin of Andronovo stockbreeding. 

Horses came third in the Andronovo herd. They were used not only for 
transportation purposes, for chariot harnessing, but also for meat, which is 



ECONOMY 149 

proved by fragmented bones on settlements. Horse -meat constituted 20-30% of 
Andronovo meat consumption (Tsalkin 1972b: 60). Probably in the late 
Andronovo period they learned to make koumiss (sour horse milk), for which 
goblet-shaped vessels were used (Margulan 1979: 262; Kuz'mina 1974). The 
horse was the main cult animal among the Andronovans, which is indicated by 
sculptures and representations of horses on Kazakhstan petroglyphs, and also 
horse burials or skins with the head and legs that substituted for them in rituals 
and Andronovo status burials of the Sintashta, Petrovka and Alakul' types. The 
Fedorovo tribes usually placed horse ribs and a shoulder-blade in a grave 
(Kuz'mina 1977: 28-52; 1980c: 32, 1994; Gening 1977). 

On the basis of horse bone measurements made by V. I. Gromova (Sal'nikov 
1951b: 124) and V. I. Tsalkin (1972b: 74-77) it has been established that the 
Andronovans bred three breeds of horses: small, up to 128-136cm at the withers; 
average and tall, up to 136-152cm high, weighing 350kg, thin-legged and semi- 
thin-legged. In addition, for the first time in Eurasia we have very large horses 
(152- 160cm), thin and semi-thin legged and distinctly graceful: these were the 
horses that were placed in elite burials (Sal'nikov 1951b: 123). The modern 
Akhaltekin breed goes back to these Andronovo horses and, ultimately, all the 
high bred horses of the world: from the elite horses of Andronovo come the 
famous Nisa horses that were bred in antiquity in Central Asia. They are the 
ancestors of the Akhaltekin horses; after the occupation of Central Asia by the 
Arabs in the Middle Ages, the Arabs took Nisa horses to Arabia, and from there 
they reached England, where English race horses were bred (Belonogov 1957; 
Barmintsev 1958). The Fedorovo horses of the Urals and eastern Kazakhstan 
differed from the Alakul' by their small height (Sal'nikov 1967: 328; Chernikov 
1960: 64; Akhinzhanov et al. 1992). Horse-breeding, reaching its climax among 
the Andronovans, developed in the south Russian steppes beginning with the 4th 
millennium BC, while at the same time in south Turkmenistan reliable recovery 
of horse bones and images of horses appear only in the Namazga VI period 
(Kuz'mina 1980c: 27-29; 1981a; 1981b). The unusually high level of horse- 
breeding constitutes the most important peculiarity of Andronovo stockbreeding 
and helps to distinguish the role of the Andronovans in Eurasian history. 

Another important feature of the Andronovo economy was the breeding of 
the Bactrian camel. Camel bones are found at the settlements of Arkaim, Atasu, 
Ust'-Narym, Petrovka 2, and Alekseevka (Tsalkin 1972b: 77; Margulan 1979: 
258, 259; Makarov 1978: 136, Lavrushin et al. 1999: 101). Ritual burials of a 
camel foal and camel bones in Aksu-Ayuly (enclosures 3, 4, 5), camel skulls 
from Telzhan-Kuzeu and near a hearth in the settlement of Milykuduk, bones in 
graves 2 and 5 in Begazy, a clay statuette on the settlement of Ushkatta, a stone 
head from the Irtysh (Fig. 55: 8-9) and numerous representations of camels on 
petroglyphs in Kazakhstan and Central Asia demonstrate the cult significance of 
this animal (Figs. 34: 10, 55: 7-9; Kuz'mina 1963a: 38-44; 1980c: 29-30; Margu- 
lan et al. 1966: 262; Margulan 1979: 258, 259; Kadyrbaev and Mar'yashev 
1977: 162-170). In the second half of the 2nd millennium BC camel-breeding 
was known in Eurasia apart from Andronovans only among Central Asian tribes 
(Itina 1977: 138, 185-190; Kuz'mina 1980: 30). 

Finally, the Andronovans bred dogs. They were not used for food, but they 
were kept as guard-dogs. Two breeds were known: a small husky type and a 



150 CHAPTER NINE 

larger guard-dog (Krivtsova-Grakova 1948: 102; Sal'nikov 1951b: 124). Dog 
burials at cemeteries and in special graves suggest an animal cult. 

The actual composition of the Andronovo herd is reconstructed on the basis 
of percentage ratio of individuals with corrections, introduced by M. P. Gryaz- 
nov (1957: 23) and Yu. A. Krasnov, taking into account the greater fertility of 
sheep (the composition for Sargary is the following: cow - 41%, horse - 37%, 
sheep - 22% (Zdanovich, S. 1979: 19). Taking into account the dead weight of 
individuals of different species it can be seen that in terms of meat consumption 
among the Andronovans beef constituted 60-70%, horse-flesh 20-30%, and 
mutton - 10% (Tsalkin 1972b: 80). 

With respect to both structure and breeds, Andronovo stockbreeding was 
structurally similar to East European, especially the Timber-grave culture (Table 
6) and it is of East European origin. It displays a clear steppe character. Its 
peculiarity is the complete absence of the pig and highly developed sheep- 
breeding that marks the Andronovo culture off from other Eneolithic and Bronze 
Age cultures of the Eurasian steppes (Tsalkin 1964: 27, 29; 1972b: 80; Kosintsev 
1989: 101; Akhinzhanov et al. 1992); in addition, it had specialized horse- 
breeding and camel-breeding. 

Horse Pig 

23.1 36.1 

16.3 14.8 

21.5 7.9 

17.6 13.1 
14.8 20.3 

16.4 10.8 
16.7 

(according to Tsalkin 1964, table 1) 
Table 6: Correlation of the MNI of domestic animals species in various archaeological 
cultures of the second half of the 2 nd millennium BC 
T-G = Timber-grave, A = Abashevo 

1. In V. I. Tsalkin's table, pig constitutes 0.4% of the domestic animals for the Andronovo 
culture but this figure was based on the settlement of Zamaraevo which is now assigned to 
the Cherkaskul' culture which did engage in pig-keeping. Later V. I. Tsalkin (1972: 79) 
stated that the Andronovans had no pigs. 

In the early and developed stages of the Andronovo culture the economy was 
settled which is shown by the long duration of settlements with very large 
houses. In the winter cattle were kept in the settlement. Andronovans stalled 
their cattle which is indicated by special fences and household annexes at 
Alekseevka and other settlements. In many Andronovo houses there is an inner 
area not so deep as the rest of the structure and containing a floor covered with 
organic waste. It is highly probable that this part of the house was employed for 
cattle keeping, especially young animals in the winter time for which there are 
numerous ethnographic parallels. Cattle fodder was hay and branches cut with 
sickle -scythes and Sosnova-Maza choppers. The stalling of cattle was a great 
achievement of the Andronovans, who considerably surpassed Western Europe 
where this system appeared only in the Iron Age. 



Culture 


Cattle 


Ovicaprids 


Pre-Kazan' 


26.9 


13.9 


Noua 


45.3 


23.6 


Pontic T-G 


46.6 


24.0 


Don T-G and A 


50.0 


19.3 


Ural T-G and A 


40.1 


24.8 


Middle Volga TG 


40.5 


32.3 


Andronovo 1 


36.7 


46.2 



ECONOMY 151 

A balance between meat and dairy products in stock-breeding guarantees a 
stable economic basis for society which leads to population growth and 
demographic expansion. In a meat-dairy economy the cattle return to the 
settlement for milking every evening. But this system limits the possible area of 
pasturage and strictly limits the number of beasts in a herd as the stocking rate on 
the Eurasian steppe is no more than six-seven head of large -hoofed animals per 
one square km of pasture. 

Human activity finds itself in a contradiction with its limited natural 
resources. In the first place, constant population growth requires growth in the 
amount of food, i.e., livestock. In the second place, even employing the most 
effective system of pasturage with the herd regularly foddered several kilometers 
round the settlement, in 20-25 years there will be overexploitation and the 
productivity of the pasture will diminish to two to four times lower. It requires 
50 years for regeneration. 

These factors conditioned the necessity of moving Timber-grave and 
Andronovo settlements to a new place, several dozens of kilometers every 20-25 
years, which explains the relatively small number of burials relating to a single 
generation in the cemeteries. 

Demographic analysis of the Elenovka micro-district involved a comparison 
of house area on the settlement with the number of burials in the associated 
cemetery (taking into account an average life expectancy in Andronovo society 
of 30 years according to V. P. Alekseev). This showed that the settlement and the 
adjacent cemetery functioned from 25 to 50 years (Kuz'mina 1974c; Evdokimov 
1984: 17). 

Settlements with mixed layers suggest several occupations at a convenient 
place to which the community returned frequently after 25-50 years. 

The gradual expansion of territory was stimulated by constant pressure from 
the surplus population. It explains a large number of monuments of the 15th- 
13th centuries BC in the steppe and the archaeologically well-traced expansion 
of the Timber-grave and Andronovo cultures that gradually mastered the whole 
Eurasian steppe. By the 13th century BC the extensive development of this 
mixed economy had come to an end. 

Intensification could now be achieved only by a transition to the new semi- 
nomadic economy and the development of new pastures. A more progressive 
system of stockbreeding appears, when the main population lived in long-term 
settlements and herds were driven to pastures, gradually moving from one 
pasture to another, which yields unlimited possibilities for increasing livestock. 
In conditions of extensive cattle-breeding seasonal pasturage is the most rational 
land-using system, that is why the transition to mountain and desert pasturing of 
cattle was a progressive phenomenon. This new system of economy was 
established among the Andronovans in the 13th— 12th centuries BC in the Final 
Bronze Age (Gryaznov 1957: 25-26; Margulan et al. 1966: 263; Margulan 1979: 
261; IK SSR: 107). However, it certainly originated much earlier. During the 
whole Bronze Age conditions for nomadic cattle-breeding gradually developed 
within Andronovo society. 

On the basis of herd composition and the topography of Eneolithic and 
Bronze Age sites, N. Ya. Merpert (1968: 41-43, 51) and V. P. Shilov (1970: 18- 
25; 1975: 5-15) established the important fact that there was a non-uniform 



152 



CHAPTER NINE 



development of various tribes and an early transition to mobile forms of stock- 
breeding in some semi-desert areas that were not suitable for a mixed economy. 
Among the Andronovans the same area specialization is noted (Map 13). The 
composition of the herd is the same along the entire territory of the Andronovo 
culture, but the relative proportion of species varies somewhat according to 
ecological conditions (Table 7). 



Region Site 


Cattle 


Sheep 


Horse 


Middle Urals 








Kamyshnoe 1 


55.6 


33.3 


11.1 


Kamyshnoe 2 


55.8 


20.9 


16.3 


Novo-Burino 


52.4 


33.3 


14.3 


Kipel' 2 


36.7 


53.3 


10.0 


Zamaraevo 3 


41.8 


41,7 


16.6 


average 


42.8 


44.5 


12.7 


South Urals 4 








Karabutak 


41.1 


35.3 


20.6 


Ushkatta 


29.2 


52.3 


18.5 


Kupukhta 


40.0 


50.0 


10.0 


Shandasha 


25.5 


64.3 


10.0 


average 


33.9 


50.5 


14.8 


Tobol 








Alekseevka 


27.9 


53.5 


18.6 


Alekseevka 1969 


36.4 


27.2 


36.4 


average 


32.2 


40.3 


27.5 


N. Kazakhstan 








Petrovka II 


42.5 


38.3 


14.2 


Novonikol'skoe 


37.4 


36.0 


19.2 


average 


39.9 


37.1 


16.7 


Sargary 


33.5 


36.2 


30.3 


Petrovka IV 5 


36.4 


34.6 


30.8 


average 


34.9 


35.6 


30.5 


regional average 


37.4 


36.5 


23.6 


C. Kazakhstan 








Buguly 6 


20.0(21.5) 


63.3 (70.0) 


13.3(8.5) 


Karkaralinsk 


30.0 (37.2) 


45.0(36.8) 


25.0(26.0) 


Shortandy-Bulak 


23.1 (21.3) 


80.8 (52.8) 


30.8 (25.9) 


Ulutau 


(33.1) 


(30.1) 


(36.8) 


Suukbulak 


(33.6) 


(42.2) 


(24.2) 


Average 


26.0 (29.4) 


63.0 (46.4) 


23.0 (24.2) 


E. Kazakhstan 7 








Kanay 


37.4 


32.0 


9.5 


Ust'-Narym 


26.4 


37.7 


7.5 


Malokrasnoyarka 


17.4 


39.1 


15.2 


Trushnikovo 


22.5 


31.4 


18.7 


average of all 








domestic animals 


38.9 


44.3 


16.8 



Table 7: Correlation of the MNI of the main domestic animals in different ecological 
zones 1 



ECONOMY 153 

/. As 100% comprises all domestic animals bones, including dog and camel, the sum in 
the table is sometimes lower than 100%. 

2. The closest numbers for KipeV, obtained by V. I. Gromova, are: cattle - 2.3%, sheep - 
47%>, horse - 8.7%. The table does not include the data for wild and domestic animals 
together. 

3. Part of the Zamaraevo materials must be assigned to the Cherkaskul' culture. 

4. Excluding the settlement ofBaytu with its small amount of material. 

5. The calculations ofS. Ya. Zdanovich (1979: 18), from which the data for Petrovka IV 
are taken, must be wrong as the sum exceeds 100%. 

6. Per cent of bones taking into account other years of excavation which is shown in 
brackets. The number of individuals was not determined. 

7. For eastern Kazakhstan the 100% includes all animal remains, including wild; the 
average here includes only the domestic species defined by V. I. Tsalkin. 

Bones of all animals, including wild ones, comprise the 100% listed. 
Unfortunately, the material is fragmentary; there are few analyses from Siberia 
due to insufficient research of settlements. According to the settlements of 
Shlyapova, Klyuchi, and Bateni the species composition was the same with cattle 
dominating the herd (Gryaznov 1957: 23; Maksimenkov 1978: 48, 51). Data 
from P. A. Kosintsev (1989: table 1) about Trans-Urals settlements and A. V. 
Galchenko (1990, table 1) for the Altai, that are reported without employing the 
MNI, are not included in the table: 



Site 


Cattle 


Ovicapr. 


Horse 


Duvanskoe 


17 


88 


64 


Cheremukhovy Kust 


106 


61 


28 


Sukhrino 3 


98 


79 


30 


Kambulat 


21 


18 


8 


Bol'shoy Log 


32 


52 


35 


Firsovo 15 


13 


9 


9 


Pereezd 


30 


37 


8 



Table 8: Number of bones 

Data about other settlements are not reliable because sampling was unsystematic. The 
data on the age of animals from P. A. Kosintsev (1989) differ from the ones published by 
V. I. Tsalkin. 

On early sites of the forest-steppe zone of the Urals and northern Kazakhstan the 
percentage of cattle ranges between 37% and 52%, averaging 40%, while sheep 
makes up only 37-44%. In contrast, in the steppe zone, in the southern Urals and 
in central Kazakhstan there is the opposite relationship of these species: cattle 
constitute only 26-34%, while sheep increases sharply to 50-63%, reaching on 
some settlements record numbers for Bronze Age sites of the Eurasian steppes: 
64% at Shandasha, 63% at Buguly and even 81% at Shortandy-Bulak (Tsalkin 
1972b: 80, table 15; Makarova 1974: 206, table, 205; 1980: 142-148, tables 1, 2; 
Akhinzhanov et al. 1992: table). 

The composition of the herd varied considerably over time. Across the whole 
Andronovo region there was a certain tendency to reduce the number of cattle, 
although the number of sheep in the forest-steppe areas changed insignificantly. 



154 CHAPTER NINE 

However, the role of the horse grew everywhere: in the Middle Urals from 10% 
(Kipel') to 16% (Zamaraevo), in eastern Kazakhstan from 7-9% to 15-18% 
(Malokrasnoyarka, Trushnikovo), in northern Kazakhstan from 14% (Petrovka 
II) to 31% (Petrovka 4), in central Kazakhstan from 24-25% (Suukbulak, Karka- 
ralinsk) reaching a maximum of 31% (Shortandy-Bulak) and 37% (Ulutau) and 
36.4% on the settlement of Alekseevka on the Tobol. In the Final Bronze Age 
the Andronovo culture became mainly horse-breeding. The increasing role of the 
horse in the transition to semi-nomadic pastoralism is especially evident; the 
herd composition remained the same, but the role of the horse grew, because 
herds were driven to remote pastures by mounted herdsmen. Also, horses can 
obtain food under snow, breaking it with the hoof. Horses eat the upper part of 
the stalks, and sheep, following them, eat the lower part of the grass; this is why 
both animal species were herded in the winter together. 

Semi-nomadic pastoralism, originating within Andronovo society, was the 
main basis for the transition to a full nomadic economy. Although it did not 
become a mass phenomenon in the Andronovo period, in some districts this form 
appeared already in Bronze Age. Nomadic society is a "society in which the 
main source of supporting life was realized by means of extensive pasture cattle- 
breeding economy" (Markov 1973: 103; 1976). S. I, Rudenko (1961) singled out 
several types of nomadism: meridional, desert, vertical, etc. The Andronovans 
were the first among the steppe tribes of Eurasia who mastered the vertical type 
of nomadism, where herds are driven away to rich mountain pastures in the 
spring. This is proved by the topography of the cemeteries that are situated near 
passes high in the mountains: in the Pamirs at Kokuybel'su and Kyzyl-Rabat 
(Litvinsky 1972: 16, 17, 29), on Chirchik in Aurakhmat and Iskander (Voronets 
1948; Kuz'mina 1966: 71-72), in Semirech'e at Tash-Tyube, Tash-Bashat, 
Dzhazy-Kechu (Kozhemyako 1960; Kuz'mina 1970; Kozhomberdiev and Galo- 
chkina 1972; Galochkina 1977) and in the Tian-Shan, where at a height 3000m 
above sea level, there is the highest mountain cemetery of Central Asia — Arpa 
(Bernshtam 1952: 19-20). 

According to G. E. Markov (1973: 103; 1976) pastoralists' acquaintance 
with high mountains makes it possible to introduce the term 'mountain-steppe 
Bronze Age' instead of 'steppe' Bronze Age. The usage of such a term seems 
unjustified. In the first place, steppe people used mountain pastures only in the 
summer; in the second place, and it is the main thing, the incorporation of moun- 
tain pasturage was a process that required many centuries and was formed in the 
steppe; that is why the Andronovo culture is better labeled a 'steppe' culture. 

The Andronovans already knew the desert type of nomadism. Evidence of 
this are the numerous sites in the sandy regions of the Caspian and Aral seas 
(Fig. 44; Formozov 1947; 1951b; Vinogradov 1959; Vinogradov et al. 1973; 
Kuz'mina 1963c; 1976b; Glikman and Melent'ev 1968; AO 1968: 147-149), in 
Khorezm (Itina 1977: 79-82, 104-109, 119-121) and in southern Turkmenia 
(Masson 1959: 27, 116; Kuz'mina 1963c; Kuz'mina and Lyapin 1984: 6-22; 
Sarianidi 1975; 1978: 549). M. A. Itina (1977) stated that short term Andronovan 
settlements cover the Tazabagyab channels and houses, i.e., they were 
abandoned during the drying of the river-beds and were connected with spring 
migrations, when delta channels were filled with water and the desert started to 
bloom. Temporary round surface houses have been found on these sites. This is a 



ECONOMY 155 

type of frame house, the predecessor of the yurt of modern nomads, that 
originated among the Andronovans and Tazabagyab people in the Final Bronze 
Age. The invention of this house in Eurasia that is ideally adjusted to the 
conditions of nomadic life was a considerable innovation in their culture 
presupposing the possibility of a transition to a nomadic way of life. 

The distribution of Andronovo and Timber-grave sites (Map 4) shows that 
migration from north to south was along the Ust'-Urt break, rich in springs and 
along the areas where underground waters reach the surface which simplifies 
well digging (Formozov 1959a: 88; Vinogradov et al. 1973: 102). Consequently, 
the invention of the well permitted the mastery of the desert by the Andronovans 
during the flowering of their culture. A comparison of the map of Andronovo 
desert sites with a map of the seasonal migrations of nomads of the Early Iron 
Age and Middle Ages shows that they coincide completely. It suggests that the 
migration routes through the deserts were first found by the Andronovans and 
that they remained unchanged for more than three millennia (Vinogradov et al. 
1973: 102-103; Kuz'mina and Lyapin 1983, Vaynberg 2000). 

So the Andronovo sites give striking evidence for the transition of a part of 
the steppe tribes to dynamic nomadism already in the Bronze Age. This new, 
more progressive form of economy led to a harmonization of the needs of society 
and the possibilities of different ecological niches. Among all the cultures of the 
Eurasian Bronze Age, the Andronovo culture is characterized by a herd 
composition most suitable for the transition to a mobile economy. Here we find: 
1) the complete absence of pig; 2) immensely large percentage of sheep which 
can find fodder under snow in winter (after horses); 3) numerous horses capable 
of obtaining food under snow, with the number of horses growing in the Final 
Bronze Age. Within the Andronovo economy for the first time was a number of 
innovations that provided an opportunity for the transition to steppe nomadism: 
1) the invention of deep wells for obtaining water in the desert; 2) the invention 
of a light -frame transportable house suitable for the nomadic way of life; 3) the 
development of wheeled transport using bullocks and specially bred heavy 
draught-horses; 4) the use of the Bactrian camel for the first time in the steppe; 
5) the appearance of horsemen in the 12th-9th centuries BC documented by rod- 
like cheek-piece types; 6) the invention of cheese providing long term food 
products from cattle-breeding; and 7) the development of a proper orientation on 
the steppe with the custom of seasonally changing one's place of habitation. 

Without these factors distant migration in the steppe, deserts and high 
mountains would have been impossible. Ecological factors were probably 
decisive. The steppe has favorable conditions for stock-breeding: in a short 
period of time livestock increases and there is population pressure which requires 
division of pasturage and forces part of the population to migrate. In other 
periods there are unfavorable conditions, epidemics that also lead to migration, 
forcing people to leave an area unsuitable for their traditional economy. 
Population density grew in some regions in the Final Bronze Age with the 
appearance of very large settlements (Evdokimov 1984). At the same time the 
cultural layer of some settlements was covered by river alluviums (Khabdulina 
and Zdanovich 1984). According to some evidence from the 12th— 10th centuries 
BC the steppe climate became more severe and damp. A complex of these 
factors contributed to the transition of the Andronovans to a semi -nomadic and 



156 CHAPTER NINE 

then nomadic economy. The reason for this transition was in the very nature of 
extensive cattle-breeding, which demanded an increase in livestock and an 
expansion of pasturage. The northern steppe territory is limited by the taiga zone 
which is unsuitable for stock-breeding. So after mastering the Eurasian steppe 
latitudinally, which was completed in the third quarter of the 2nd millennium 
BC, new areas of economic significance could be sought only in the south, and 
this defined the next direction of migration. 

In the farming cultures of south Central Asia and Iran there were neither the 
preconditions nor reasons for migration into the steppes, which were unsuitable 
for their traditional irrigation-farming economy. Desert survey has shown that 
sherds of wheel-made ceramics of the Anau type are not found further than the 
first ridge of sand dunes. There are no archaeological traces of a migration of the 
farming cultures to the north in the 3rd-2nd millennia BC. 

A study of the economic types of the Eurasian steppe tribes in the Bronze 
Age and in the Eneolithic (Merpert 1968: 41-43, 51; Shilov 1970: 18-25; 1975; 
Kuz'mina 1981c; 1986c; 1996) as well as the analysis of the specifics of Andro- 
novo cattle-breeding elucidates the history of pastoral nomadism in Eurasia. This 
has caused sharp debates which have lasted for years between A. J. Toynbee 
(1935: 404) who assigned the beginning of nomadism to the 4th-3rd millennia 
BC and who overestimated the role of nomads in Old World history, and his 
rivals, who thought that the transition to nomadic pastoralism only happened on 
the border of the 2nd-lst millennium BC (Khazanov 1973: 5-10) or in the 8th- 
7th centuries BC and resembled a sudden jump (Gryaznov 1955; Markov 1973: 
109). 

In reality the development of the pastoral economy was a regular historic 
process preconditioned by population demand to adjust its economy to the 
ecological conditions of the steppe. In spite of widespread opinion the nomadic 
way of life was already established in the Final Bronze Age. But at the crisis 
period experienced by Andronovo society at the end of the 2nd millennium BC 
when a large part of the steppe population moved from a mixed settled economy 
to pastoral nomadism, another part of the Andronovo tribes looked for a way out 
by intensifying farming and probably in economic specialization, which is shown 
by irrigation and water control constructions, the modification of the old and the 
search for new types of farming tools (specialized plate-like sickles with folded- 
over sleeves, adzes with edges, celt-spades). This conclusion contradicts the 
generally accepted mass transition to nomadism (Margulan et al. 1966: 261; IK 
SSR: 107, 112; Markov 1973: 105) but it is confirmed by evidence about 
Andronovo house -building: while everywhere instead of large long-term houses 
with stone and wood constructions there appear light-frame constructions and 
houses of the proto-yurt type, in some districts stone architecture reaches its 
peak. Such ecological niches with settled life and intensive farming were the 
regions, rich in mines, in central and eastern Kazakhstan and the fertile foothills 
of Semirech'e and the valleys of Central Asia. 

In Central Asia the settling of steppe pastoralists and their transition to 
farming is especially clear, and can be seen in the related farming cultures, Chust 
in Fergana, Burgulyuk in the Tashkent oasis, which were formed in the steppe 
zone. The decisive role of the steppe component in the Chust culture is 
convincingly shown by Yu. A. Zadneprovsky (1962). The same genesis is seen 



ECONOMY 157 

for the Bishkent (Vakhsh) culture in northern Bactria (Mandel'shtam 1968; 
Kuz'mina 1972a). 

At the end of the 2nd millennium BC there were three types of cultures 
across Central Asia: 

1) primordially farming cultures, genetically connected with the ancient 
farming cultures of Iran and Central Asia, which continued in the oases of 
southern Turkmenia in the territory of future Parthia and Margiana and which 
advanced into Bactria; 

2) farming cultures, created as the result of the settlement of pastoral tribes, 
genetically connected with Eurasian steppes populations, developing in the 
steppe regions of Central Asia, in Khorezm, Fergana, the Tashkent oasis and 
Bactria; 

3) pastoral cultures, created as the result of the transition to nomadic 
stockbreeding by pastoral tribes, that were genetically connected with the 
Eurasian steppe population and related to the farming culture of type 2 that had 
undergone sedentarization (Kuz'mina 1972a: 144). 

Both related groups of steppe populations migrated to the south and 
assimilated the originally farming cultures of south Turkmenia and Bactria, that 
were genetically alien to them. This resulted in the formation of a series of 
'barbarian' cultures of the border of the 2nd/lst millennium BC such as the 
Yashilli in south Turkmenia, Yaz 1 in Margiana, Kuchuk in northern Bactria and 
Tillya in southern Bactria. 

The formation of the languages and cultures of these groups from the steppes 
and Central Asia at the end of the Bronze Age became the genetic basis of the 
further development of Central Eurasia in the early Iron Age when Iranian- 
speaking tribes performed in the world historical arena. These tribes were the 
nomadic Sauromatians of the Volga and the Ural, and the Saka in Kazakhstan, 
southern Siberia, and the south of Central Asia. South of Central Asia, in Iran 
and Afghanistan, they were the farmers of Khorezm, the Sogdians, Margianians, 
Bactrians, Medes, and Persians. Their language and history are known to us from 
the descriptions of Greek authors and documents of Achaemenid Empire. 

The Indo-Iranian economy 

The Indo-Iranian economy is reconstructed, first on the basis of common Indo- 
European words preserved in their languages, second from the terms present in 
the earliest Vedic literature, summarized by A. Macdonell and A. Keith (1958), 
and finally from the ancient texts analyzed in many works, that reconstruct 
Aryan society and way of life (Zimmer 1879; Bhargava 1971; Ghurye 1979; 
Srinivas 1982, Sharma 1983, Pandey 1965). Although the work of some Indian 
scholars suffers from attributing to the Vedic period a relatively modern form of 
economy and social system, the conclusions of W. Rau (1975; 1983) and T. Ya. 
Elizarenkova (1972; 1989; Elizarenkova and Toporov 1995) are extremely 
important and are based on a thorough analysis of the sources. 

Works by W. Geiger (1882), Ch. Bartholomae (1904), M. D. Dhalla (1922), 
M. M. D'yakonov (1954), V. I. Abaev (1956), V. A. Livshits (1963; 1998), R. 
Frye (1962; 1972; 1976; 1996; 2002), and M. Schwartz (1985) are of the utmost 
importance for the reconstruction of the Avestan economy and society. 



158 CHAPTER NINE 

According to linguistic data the Indo-Iranian economy was mixed farming 
and stockraising, with stockraising dominant. I, M. Steblin-Kamensky (1982) 
showed that there is often a shift of meanings in the names of cultivated plants. 
A whole semantic bundle is connected with the Indo-European word for 'grain', 
initially in early Indian and Avestan 'barley' as in the other Indo-European 
languages. The names of wheat in some Pamir languages go back to it, and in 
Scythian and some other Pamir languages it indicates 'millet', also Avestan 
'plow land', 'crop', Afghan 'plow'. The word 'grain' is an ancient borrowing 
into Finno-Ugrian from Indo-Iranian (Steblin-Kamensky 1982: 22-26). The word 
'barley' is an Indo-European term (Gamkrelidze and Ivanov 1984: 655). Barley 
is mentioned in the Rigveda, but wheat and rice are absent there (Elizarenkova 
and Toporov 1995: 502). Common Indo-European are words for 'chaff, 'awn', 
'grinding' ('millstone'), as well as a semantic bundle originating from the verb 
'to crush': 'pestle', 'flour', 'oat flour' (Schrader 1913; Steblin-Kamensky 1982: 
27, 48). Common Indo-European is 'millet' that is probably close to Greek; the 
word for 'rye' has no Indo-European roots; the early Iranian term denotes 
'barley-like' (Steblin-Kamensky 1982: 30, 33). 

This linguistic evidence corresponds to literary testimony. Yima received a 
divine gift — a plow. In the Rigveda and Avesta tilling with a pair of bulls 
harnessed to a plow is mentioned as well as barley, hay, stone mortar and pestle, 
copper sickle and axe for slaughtering cattle. Public wells are also known, but 
reference to irrigation constructions is doubtful (D'yakonov, M. 1954; Elizaren- 
kova 1972: 38; 1989: 449, 451; Elizarenkova and Toporov 1995: 512, 513, 519, 
Dhalla 1922: 142; Rau 1975; 1983). 

Cattle-raising was the basis of the Indo-Iranian economy. In Vedic texts there 
are about twenty words to denote a horse, ten for a bull, fifteen for a cow; these 
have numerous epithets (Elizarenkova 1982: 39, 42, 47; Elizarenkova and 
Toporov 1995: 502, 503), and are constantly compared to gods. Sheep, goat, and 
the Bactrian camel are also known. Domestic animal names in the Indo-Iranian 
languages have secure Indo-European cognates. There are several terms denoting 
'bull', 'cow', 'cattle', words for 'beef, 'udder', 'to milk', 'milk', 'butter', and 
also an Indo-Iranian -Baltic term to denote 'milk' and an Indo-Iranian-Greek- 
Slavonic term to denote 'curds' (cheese), several Vedic terms to denote dairy 
products, showing the special development of dairy cattle-raising among the 
Indo-Iranians (Schrader 1913; Abaev 1979 III: 155-156, 319; Gamkrelidze and 
Ivanov 1984: 563-573; Lehmann 1973; Elizarenkova and Toporov 1995: 519). 

Of common Indo-European origin are also the Indian and Iranian names 
'sheep', 'fleece', and 'wool' (Gamkrelidze and Ivanov 1984: 577-580). V. A. 
Livshits has made some important contributions. An ancient term for 'cattle' was 
recorded in the Avesta and was then later attributed to 'sheep' in the Iranian lan- 
guages; Yima's sacrifice of cattle (Yasna 32.8) was replaced by a sheep sacri- 
fice. These facts indicate that the rise of sheep-raising in Iranian society occurred 
after the collapse of Indo-Iranian unity. The etymology of the early Persian word 
'palace' (used for instance for Persepolis) denoted 'fence', 'pen', 'sheep-fold'. 

Goat-raising was not significant among the Indo-Iranians. The words 'goat' 
and 'ram' have only regional correspondences. 

The exclusive role of horse-raising, the presence of camel-raising and the 
complete absence of the pig make up the characteristic features of Indo-Iranian 



ECONOMY 159 

stock-raising. The Indo-European term 'piglet' has been preserved only in 
Khotanese Saka (Burrow 1976: 1 1), and the other Indo-Iranian names for the pig 
refer to the wild boar. 

E. Benveniste (1970: 307-320) showed that the common Indo-European term 
'cattle' denoted 'movable property'. In Indo-Iranian languages the word 'war' 
and 'booty' literally denote 'cattle stealing' (Abaev 1979 III: 122-3, 149, 171). 

Evidence for cult practice is an important source for studying Indo-Iranian 
stockbreeding. It has been stated that an animal's value was defined by its place 
in society. According to the Rigveda and Avesta horse sacrifice occupied the first 
place, the bull was second and sheep third (Ivanov 1974; Schmidt 1980). Dog 
was of considerable significance among the Indo -Iranians, especially the 
Iranians: it is a friend of man, sometimes follows him after death; it is protector 
of every family and its cattle, that is why the dog is honored in the Avesta; to kill 
it was considered a grave sin (Dhalla 1922: 157, 178). 

The predominant role of cattle-raising among the Indo-Iranians is revealed in 
requests to the gods and in the constant epithets of the gods. In Rigvedic hymns 
(2.1.5; 3.1.23; 3.15.7) Agni calls a cow as 'giving wonderful wealth', 'much 
promising reward'; Indra (3.31.4; 7.18.22) is called 'ruler of golden horses', 
'generous with oxen', 'herdsman of cows', 'lord of cows', the one whom "the 
noble cow allows to milk sweet honey"; he is prayed "to make us seizers of 
cows"; he gave people 'well milked cows, horses'; Mitra-Varuna — "wonderfully 
giving shepherds of the Universe make all milking cows in an enclosure swell" 
(5.62.2). The Asvins (1.118.2; 1.157.2; 7.71) are summoned to "make our cows 
swell from milk, fast horses frisky"; "with fat and honey sprinkle our domina- 
tion"; "oxen-givers, horse -givers, come, bringing good". Visnu is prayed to 
"provide us with wealth consisting of horses" (7.100.2), Agni is asked to bring in 
cattle (5.2.5,12), take belongings from the enemy; to give wealth of fast horses 
(5.6.10), "wide stream of milk, basis of wealth" (5.15.5), "you are... the giver of 
reward, made of cows" (5.23.2). Indra is "owner of dun horses", "herds in front 
of well-milked cows", "as a shepherd herds of cows" (5.31.2,3); he "did much 
for a cow, fighting for pastures" (5.33.4). Mitra and Varuna are asked to give as 
reward the winning of cattle (5.41.1), they "increase plants, make cows swell 
from milk, pour out rain" (5.62.3). Gods are asked to "bring wealth, consisting of 
harnessed horses" (5.41.5). The Maruts provided hundreds of cows, an honorary 
gift of cows, of horses (5.52.17). Parjanya is asked: "let there be a good 
watering-place for cows! You made deserts easy to pass" (5.83.8,10). Ptisan, the 
cattle protector (1.42), is addressed with the prayer to preserve one on the road 
from wolf and robber, to create easy-traveled roads, "to bring to good pastures"; 
in one hymn (1.22.4) there is an appeal to all the gods: 

Let us be sought ways 

For a happy journey 

Indra, Maruts, 

Pusan, Bhaga, worthy of worship! 

The latter hymns make it possible for us to accept the opinion that the Vedic 
Aryans were not settled but mobile cattle-keepers (Rau 1983; Elizarenkova and 
Toporov 1995: 489, 490). They spent half a year in settlements from the sowing 
of barley to its harvest, for half a year their house was a wagon and proto-yurt, 



160 CHAPTER NINE 

they conducted wars using their chariots to conquer new lands and cattle. There 
is "no hints of city life in the Avesta" (Livshits 1998: 221). 

In the earlier parts of the Avesta the Iranian economy is also presented as 
cattle-breeding. Cattle is the main property (Livshits 1998: 218, 222), the word 
for 'house' (nmana) denotes also a 'room for cattle, cattle-shed' (Videvdat 2.23; 
16.14), the word 'property' means 'living existence', hence 'means of existence'. 
'Houses rich with cattle' are mentioned (Yast 17.8). Zarathustra prays for giving 
cattle to the honest cattle-breeder, for increase of cattle (Yasna 31.15; 45.9). The 
cultural hero Yima bears a constant epithet 'possessing good herds' (Yasna 9.4, 
5; Yast 5.25; 17.31; 19.31, 34). Ahura-Mazdah is prayed for "the abundant in 
pasture beautiful life" to come (Yasna 48.1 1). "For all time there are codes about 
the pastoral life, which must not be broken by anybody" (Yasna 51.14). 

The Vedas reflect the mobile way of life; in the Rigveda there are more than 
fifteen words denoting 'way - road' (Elizarenkova and Toporov 1995: 511; 
Elizarenkova 2000). In the Mihr Yast (9) Mithra has the constant epithet 'master 
of vast pastures'. In Yasts 10.28 and 13.52 it is said: "We worship Mithra, 
possessing vast pastures for cattle, Mithra who presents herds of cattle", "let him 
put in this house herds of cows". Mithra, Tistriya, Haoma (Yast 10.28, 65; Yast 
9.17, 19) are called to present cattle and horses, water pastures. The goddess 
Anahita is invoked as 'increasing herds', 'increasing property', 'increasing 
wealth' : "Let me conquer vast kingdoms, where horses are snorting, wheels are 
ringing" (ArdvTsur Yast). Cattle is the measure of wealth (Yasna 46.2; 29.12; 
Yast 5.25; 12.31; 19.31,34). 

The goddess Asi is protectress of cattle (Yast 18.4, 5). In the Mihr Yast 
(10.38) it is said about the driving away of cattle: "the cow, accustomed to 
pastures, is driven along the dusty road of captivity, dragged forward in the 
clutches of treaty-infringing men as their draught-animal; choking with tears 
they (= cows) stand, slobbering at the mouth" (Gershevitch 1959: 93). The soul 
of the cow appeals to Ahura-Mazdah to defend cattle from being stolen and 
driven away. Punishment follows for cow stealing and maltreatment of animals 
(Yast 10.38, 86). Highest virtues are "to praise Ahura-Mazdah and give fodder to 
cattle" (Yasna 35.7). 

The Indo-Iranian word for 'social man', the main producer, denotes 
'shepherd' (literally 'bringing fodder to cattle', D'yakonov M. 1961: 61, 363) 
which is related to Hittite 'shepherd' (Ivanov 1957). Vedic grama 'village, 
community' is cognate with Slavonic gromada, that initially denoted a group of 
traveling shepherds (Rau 1983, Elizarenkova and Toporov 1995: 490, 491). 

Thus, the Indo-Iranian economy was mixed farming-cattle-raising with 
herdsmen dominant; it was relatively mobile already in the Vedic period. It has 
neither analogies nor sources in the whole Indo-Near Eastern region. There the 
economic basis was irrigation farming; stock-raising played a secondary role and 
differed in species, and in the percentage or species: sheep and goat occupied the 
first place, cattle, often of the zebu breed came second, pig played an important 
role, donkey was the main transportation animal, the horse was known from the 
3rd millennium BC in Mesopotamia, but it was not used in the economy (Tsalkin 
1970b; 1972a; Masson 1964; 1976). 

In the Indo-Iranian herd cattle, on the contrary, occupied the first place, 
ovicaprids were represented mainly by sheep, and horse-raising occupied an 



ECONOMY 161 

exceptional role. Such a type of economy in the Bronze Age existed only in the 
Eurasian steppes of the Old World, where one must look for the Indo-Iranian 
homeland. 

For a more precise cultural attribution the absence of pig-breeding among the 
Indo-Iranians and their acquaintance with the Bactrian camel are decisive. Of all 
the pastoral steppe cultures these peculiarities are typical of the Andronovo 
culture, economic and cultural type (ECT) which corresponds fully with that of 
the Indo-Iranians. 

Indisputable proof of the Aryans coming into India has been demonstrated by 
F. B. J. Kuiper (1991). For the flora and fauna of the Rigveda all south Indian 
plants, animals, and birds are called by loanwords from Dravidian and Munda, 
every day objects and crafts are also borrowed, and what is more important, 
terms connected with farming: irrigation, channel, gutter, irrigated field. This 
reflects the influence of a non -Aryan farming population with a different type of 
culture (Kuiper 1991: 14, 15, 21) and the gradual adaptation of Aryan cattle- 
breeders to a different ecological zone and a different ECT. 

The historic fate of the Iranian tribes is clarified by the Gathas which, 
according to tradition, were created by Zarathustra. There is the call to reject 
pastoralism and pass over to settled farming (D'yakonov, M. 1954; Abaev 1956). 
The symbol of faith of the Zoroastrians is (Yasna 12): "I... reject stealing and 
seizure, from damage bringing and destruction of Mazda settlements", "the root 
of belief of Mazda worship... is plowing; he who plows bread, plows truth". 
This makes it possible to relate the culture of the Iranian-speaking peoples at the 
time of Zarathustra to the culture of the barbarian occupation, when steppe 
pastoralists moved from the north to the south of Central Asia and Afghanistan 
and passed over to a sedentary life and the assimilation of the ancient local 
farming population. 



CHAPTER TEN 

THE ETHNOGENESIS OF THE INDO-IRANIANS AND THE 
ETHNIC ATTRIBUTION OF THE ANDRONOVO CULTURE 

The retrospective approach 

Our analysis of material culture carried out according to the methods discussed 
earlier shows that in the 2nd millennium BC there were two large zones, Central 
Eurasian and Indo-Near Eastern, both developing independently from the 
Neolithic-Eneolithic period. According to its economic and cultural type the 
Andronovo culture belonged to the Central Eurasian zone as it culturally 
paralleled the pastoral cultures of the steppes of southern Russia, especially 
synchronous cultures such as the Timber-grave culture, which were evolving in a 
parallel way. The differences between these cultures are limited to some details 
of house construction, ceramics and burial rite which, as we suggest, are the very 
features that determine their ethnic affiliation and that justify their separation 
into two independent cultural units. The third related yet independent culture is 
the Tazabagyab, which differs from the Timber-grave and Andronovo in the 
large role of irrigation agriculture in the economy, house type and ceramics. 

Comparing the categories of Andronovo material culture with those of the 
following Indo-Iranian Sauromatians and Saka, we discern a genetic succession 
determined not by isolated categories of artifacts but a systematically and 
transparently interrelated all-encompassing typological complex. It is important 
to stress that no definite interactions with the south — from the zone of ancient 
agriculture — mitigated the transition to the Early Iron Age in the steppes. All 
elements of Saka material culture that were ideally adapted to the ecological 
conditions of the steppe had been slowly maturing during the previous 
millennium and the transition to nomadic pastoralism arose out of the Final 
Bronze Age when livestock, transport, tools, clothes and type of housing were 
established to form the preconditions for passing to the new type of 
pastoralnomadism. The military traditions of the nomads were an important part 
of their culture, emphasized by Persian as well as Greek authors, and these had 
their roots in the Bronze Age: the equestrian tradition succeeded the chariot 
combat tactics in the late 2nd millennium BC when new types of cheek-pieces 
appear to be developed in the Iron Age; the main types of early-Scythian arrows 
derive from the Andronovo- Alakul' (Kuz'mina 1985b) as well as different types 
of Saka spears, axes, adzes, knives, and sickles, which all form continuous 
typological series (Kuz'mina 1966; 1985b). 

Succession is determined not only through the elements of material culture, 
defined as belonging to an economic-cultural type, but also through those func- 
tionally non-defined but ethnically significant traditions: technologies and the 
ornamentation of ceramics. Despite the transition to mass nomadism the 
traditions of post-frame house architecture survived in the 'house of the dead', 



164 CHAPTER TEN 

the sub-kurgan chamber that was ritualized and, hence, an important ethnic 
indicator. Details of costume, particularly headdress, that served as tribal 
indicators even in ethnonyms (pointed-hat Saka) also go back to the Andronovo. 
K. F. Smirnov (1957b; 1964), K. A. Akishev (1973); Akishev and Kushaev 
(1963), B. A. Litvinsky (1972) and M. K. Kadyrbaev (1966: 408-409) estab- 
lished the succession of the burial rite, particularly the construction of kurgans, 
stone enclosures, different types of graves, and the prevailing westward 
orientation. This provides an objective foundation for establishing a genetic 
connection between the Sauromatian and Saka cultures in the 7th-6th centuries 
BC and the Andronovo culture so that an Iranian or Indo-Iranian attribution of 
the latter is well substantiated by the retrospective method. 

The evidence of material culture 

Following the method suggested in Chapter 1 we compared the material culture 
of the ancestral Indo-Iranians as reconstructed from the evidence of language and 
written testimony with the economic and cultural types of the Old World (2nd 
millennium BC); we specifically compared the Central Eurasian zone (including 
the pastoral cultures of the Eurasian steppes) and the Indo-Near Eastern zone 
(including agricultural and stock-breeding cultures). 

Already in the 19th century the analysis of the vocabulary of the Indo-Iranian 
languages, primarily the vocabulary of the oldest written sources that preserve 
Indo-Iranian tradition, the Avesta and the Vedas (primarily the Rigveda), estab- 
lished that the Indo-Iranians lived in the vast steppe with large deep rivers and 
they had a mixed economy with stock-breeding predominant (Geiger 1882; Ol- 
denberg 1894; Schrader 1901; Pedersen 1931; Marquart 1938, D'yakonov 1956; 
Ivanov and Toporov 1960; D'yakonov M. 1961; HTP 1963; Bongard-Levin and 
H'in 1969; Grantovsky 1970; 1980; 1988; Abaev 1972; Gafurov 1973; Eliza- 
renkova 1972; Boyce 1975; Burrow 1976). This is demonstrated by the fact that 
though the Indo-Iranian languages have general Indo-European names for cereals 
and the plow, they have a poorer agricultural vocabulary than other Indo- 
European languages. And, in contrast, cattle- and horse-breeding terms are 
numerous. Indo-Iranian gods have epithets such as 'master of vast pastures', 
'granting the richness of cattle', and 'splendid richness of horses'. The gods are 
constantly asked to grant rich livestock, especially horses, to protect cattle, to 
water the pastures and thus it leaves no doubt that cattle-breeding dominated the 
economy of the Indo-Iranians. 

Livestock and the relative value of the animals are defined by numerous 
descriptions of sacrifices: the highest sacrificial animal was the horse followed 
by cattle and sheep. The Sanskrit word 'war' literally means strive to capture 
cattle (Schrader 1901; 1913). V. A. Livshits noted that in the Persian language 
the sacrificed sheep is designated by the word gospand which derives from 
Iranian 'holy' + 'cow', hence, cattle did prevail in the herd of the ancient Indo- 
Iranians. Unlike other Indo-Europeans the Indo-Iranians did not raise pig, but 
they bred Bactrian camels (Kuz'mina 1963a; Bulliet 1975). 

Indo-Iranian pottery was made by hand for domestic use without the potter's 
wheel (Sinha 1969; Rau 1972) (Table 9). W. Rau showed that the earliest Indo- 
Iranians did not know a professional craft with narrow specialization; they relied 
on domestic production. They knew metallurgy and metal processing (Rau 1973; 



THE ETHNOGENESIS OF THE INDO- IRANIANS 



165 



Bailey 1975; Jule 1985), including casting in special molds followed by forging 
and sharpening, that parallels the development of Andronovo metallurgy; along- 
side bronze, stone was also employed for the manufacture of tools and weapons 
such as querns, rubbing stones, utilitarian axes and battle axes, maces, ritual 
vessels — the types well known in Andronovo complexes. 



Feature 


Zone of agricultural cultures 
of the Near East and Central 
Asia 


Zone of cattle-breeding 
cultures of eurasian steppe 


Production 


Specialist male occupation 


Non-specialist female domestic 
wares 


Consumption 


Market 


Family 


Technology 


Potter's wheel 


Hand-made 


Pyro-technology 


Special kiln 


On lire or in hearth 


Ornamentation 


Slipping, painting, unornamented 


Stamped ornament, applied-roller 



Table 9: Comparison of ceramic traditions of the 3rd-2nd millennia BC 

Chariot making was very important. The first chariot-maker was the creator-god 
Tvastar. The words 'chariot', 'shaft', 'saddle-girth', 'bridle' and many names of 
harness parts were common in the Indo-Iranian languages (Abaev 1949; 1972; 
D'yakonov 1956; Bailey 1955; 1957; 1958; Grantovsky 1970). Some elements 
of dress and weapons, including trousers and hood — the traditional dress of 
Eurasian steppe nomads — are also common (Bailey 1957). 

However, there is no evidence for a general Indo-European terminology con- 
nected with specialized crafts and trade, irrigation, cities, fortification, monu- 
mental palaces and temples, cult images, and written language. Palaces iyimana) 
where Indo-Aryan gods live in heaven, according to J. Hertel, are huge vehicles, 
'houses on wheels'. Iranians called a palace hadis, i.e., 'cattle pen', 'fold'. This 
is what the palace of Darius in Persepolis was called (Benveniste 1955). The 
common Iranian name for a dwelling place (katd) derives from the verb kan- 'to 
dig', i.e., denotes a semi-subterranian house (Benveniste 1955: 301). In the 
Avesta (5.101) the house of Anahita is described as a large post-frame house 
with a light opening and (timber) pillars. L. Renou (1939) analyzed such house 
descriptions in the Atharvaveda (3.12.1) to show that the common Indian and 
Iranian terminology and construction involved a large post-built semi-subter- 
ranian house with a gable roof (Table 10). The second type of dwelling place — 
the proto-yurt — can also be reconstructed. A word in the Middle and New 
Iranian languages for a tent or yurt-type house comes from Avestan vi-da- ('to 
make a house') (ITN: 141; Benveniste 1955: 301). These linguistic data provide 
unquestionable evidence that not only proto-Vedic Aryans but also Iranians were 
not familiar with monumental architecture in the pre-Achaeminid period. Nor did 
they know temples. Herodotus (1.131) notes that there were no temples even in 
the Achaemenid period. They are mentioned neither in Vedic texts nor in the 
Mahabharata. The term pur ('fortress') is used to designate only aboriginal 
settlements and Indra is called 'destroyer of purs' (Rigveda 4.30.20). 

Hence, the economy of the Indo-Iranians in their homeland is reconstructed 
as mixed agricultural-stockbreeding with stockbreeding predominant and with 
distinct features of the transition to pastoralism and even nomadic forms already 



166 



CHAPTER TEN 



with the Indo-Aryans. "We have neither houses nor fields... We are well in the 
world as people who indeed spend their lives on wheels" (Visnupurana 5,10,33). 



Feature 


Zone of agricultural cultures in 
the Near East and Central Asia 


Zone of stock-breeding cultures in 
the Eurasian steppes 


Type of dwelling 
place 


Surface 


Semi-subterranian post-frame 


Construction 
material 


Bricks, pakhsa, clay; rarely stone 
blocks horizontally set on clay 
mortar 


Wood, poles smeared with clay, baira; 
rarely stone slabs erected vertically 


General plan 


Small square room 9 -12 m 2 


Large rectangular dwelling place 50- 
300 m 2 


Inter-house 
connections 


Continuous living quarters of 
building system with open inner 
yard in the center 


Separate dwellings, sometimes 
connected to a household annex 


Type of covering 


Flat roof 


Gable roof, multistage-pyramid vault 


Type of heating 


Closed surface hearth, fireplace 


Open circular, subterranian hearth, fire 


Entrance 


Doorway 


Long corridor-ramp ipandus) 


Social function 


House for small family 


House for large family community 


Main settlement 
type 


Compact plan of settlement with 
continuous blocks of multi-room 
houses 


Dispersed plan of settlement with 
separate houses 



Table 10: Comparison of dwelling places in the 2nd millennium BC 

We see then, by a complex system of analysis the material culture of the 
ancestors of the Indo-Iranians cannot be assigned to the economic and cultural 
type of the Indo-Near Eastern zone where the economy was based on irrigation 
agriculture since the Eneolithic, where cattle breeding played a subsidiary role to 
mostly sheep and goat and animals such as the pig and donkeys, which were 
alien to the Indo-Iranians, were reared; the animal common to Indo-Iranian 
culture, the horse, was missing; where there were fortified cities, palaces, 
temples, houses with small rooms and an inner yard of the Indo-Near Eastern 
type; where craft specialization was developed that worked not for the family but 
for the market; where different types of wheel-made ceramics and a different 
costume or dress were used. 

This is supported by the conclusions by F.B.J. Kuiper (1991) who demon- 
strated that terms connected with the flora and fauna of India, the potter's craft, 
irrigation and agriculture are borrowed in the Rigveda from aboriginal languages. 

The culture of stock-breeding communities in the Near East cannot be 
labeled Indo-Iranian either, as the detached groups of stockbreeders were few, 
poor, disconnected, huddled in the deserts and mountains at the margins of fertile 
areas and, judging from the written sources of the 2nd millennium BC analyzed 
by H. Klengel (1967), J. R. Kupper (1957) and R. Walz (1959), specialized espe- 
cially in sheep-breeding and since the 1st millennium BC started breeding 
camels (Dromedary), so their type of economy and level of culture also differed 
from that of the Indo-Iranians. 

Therefore, our analysis of the Indo-Iranian type of economy and culture 
(Table 1 1) firmly places their culture in the central Eurasian zone, most closely 



THE ETHNOGENESIS OF THE INDO- IRANIANS 



167 



related to the pastoral cultures of the Eurasian steppes: the Timber-grave and 
mainly Andronovo. 



Feature 



Agriculture 


Irrigation 




Non-irrigation 


Stock-breeding 


Horse 




Cow 




Sheep 




Camel 




Pig 




Donkey 


Metal production 


Slaves 




Community 


Crafts 


Specialized 




Domestic 


Settlement 


Town 




Village 


Structures 


Fortification 




Temple-palace 


Writing 




Seals 




Costume 


Cap 




Trousers 




Boots 




Sandals 



Indo-Near Eastern 
Agriculture Stock-breeding 
+ 



Eurasian Steppe 
Andronovo Saka 



Table 1 1 : Comparison of economic-cultural types of the 3rd-2nd millennia BC and the 
Saka (1. only in Anau) 

It is probable from the linguistic data that in the middle of the 2nd millennium 
BC the Indo-Iranian language family was an aggregate of independent related 
languages and dialects that constituted a language continuum. It is possible to 
suppose that separate local variants and types of sites of the Andronovo culture 
corresponded to separate tribes who spoke certain dialects. Such an under- 
standing of the Indo-Iranian continuum allows us to attribute the Timber-grave 
culture, which was a possible ancestor of the Iranian-speaking Scythians, to 
Proto-Iranian. It seems that earlier types of Indo-Iranian ethnogenesis cannot be 
established at present with the same degree of probability. This is because the 
important cultural innovations (horse-drawn chariots, bronze alloys) that 
appeared in the Eurasian steppes and the subsequent migrations that took place 
within the zone were probably of a corporate character and involved different 
cultures (Poltavka, Catacomb, Multi-roller ceramics, Abashevo); these processes 
terminated to form two large cultural unities: the Timber -grave in the west and 
Andronovo in the eastern steppe. Conclusions concerning the languages of the 
corporate migrants are impossible without additional data. 

The hypotheses that the bearers of the Catacomb and Abashevo cultures were 
Indo-Iranian cannot be strictly proven. First, to prove this we cannot apply the 



168 CHAPTER TEN 

retrospective method that we recognize to be decisive as their direct descendants 
and their languages have not been established. Second, it involes the comparison 
of single arbitrarily chosen instead of a complex of systematically bound features 
that are a precondition for reliable ethnic reconstruction; the set of interrelated 
cultural features in general does not correspond to Indo-Iranian culture as 
specific ethnic indicators are missing. Third, these hypotheses cannot be verified 
with the help of other independent data. However, as the Poltavka, Catacomb, 
Multi-roller ceramics and Abashevo cultures probably took part in the formation 
of the Andronovo and Timber-grave cultures to some extent, so providing them 
with a Proto-Indo-Iranian identity can hardly be excluded. 

The next step in identifying an Indo-Iranian ethnos with a concrete archae- 
ological culture is to uncover those ethnic indicators that clearly distinguish 
Indo-Iranian culture from that of other Indo-Europeans. These ethnic indicators 
include the fact that of all the Indo-European peoples virtually only the Indo- 
Iranians did not raise pigs nor sacrifice them and only the Indo-Iranians raised 
Bactrian camels and had a cult of this animal. As far as I know the only Bronze 
Age culture that did not practice pig raising was the Andronovo. As for the 
Timber-grave culture, pig is present in osteological materials although there is no 
cult of the animal. The culture of all Indo-Iranians had a developed cult of the 
camel in combination with the cult of the horse (Kuz'mina 1963a; Bulliet 1975). 
In contrast, peoples in the Near East knew only the Dromedary camel that never 
was a cult animal until Assyrian times (Salonen 1955-56). All Indo-European 
peoples except for the Indo-Iranians had no traces of an ancient acquaintance 
with the camel and nor traces of its cult. The name of the camel in the Indo- 
European languages, except for Indo-Iranian, is a later borrowing of a Semitic 
gamal. In the Indo-Iranian languages the Bactrian camel is called by a general 
word — the ancient Indian ustra-, so it follows that the Indo-Iranians were acquai- 
nted with the animal after they had left the Indo-European homeland but had not 
yet separated into the Indian and Iranian branches. The cult of the Bactrian camel 
represented in Iranian mythology was formed in a zone where they continued to 
be in contact with each other. According to paleozoological and archaeological 
data the Andronovans who migrated from the south Russian steppes first became 
acquainted with the Bactrian camel in the 2nd millennium BC, and in the Andro- 
novo and Tazabagyab cultures camel-raising was established and the cult of this 
animal was formed (Kuz'mina 1963a; 1980b; Margulan et al. 1966; Itina 1977). 
Numerous Bactrian figures proving the cult of this animal are found on petro- 
glyphs in Kazakhstan and Central Asia (Kadurbaev and Mar'yashev 1977; Sher 
1978; Martynov et al. 1992, Mar'yashev and Goryachev 1998). Thus, the Andro- 
novo culture is the only culture in which the Bactrian camel, horse, ox, and sheep 
cults are combined, and there is no pig cult. In this culture horse-drawn chariots 
spread early, there was a chariot cult and elite chariot-warriors, the social strata 
of society was defined; the fire cult was developed (including a hearth and ash 
cult); the dead were buried in graves under a kurgan with a fence according to 
cremation or inhumation ritual; the economy involved mixed farming with stock- 
raising dominant. In other words the economy, everyday life, social system, 
ritual and beliefs of the Andronovans corresponds completely to the picture that 
is reconstructed for the Indo-Iranians according to the evidence of language, 
which leads to the conclusion that the Andronovans spoke Indo-Iranian. 



CHAPTER ELEVEN 
VERIFICATION OF THE HYPOTHESIS 

Anthropological evidence 

Anthropology provides an absolutely independent and extremely useful source of 
information for testing a hypothesis. The close relationship between the western 
Andronovo population and the Timber-grave people in the contact zone 
stretching from the Volga to central Kazakhstan is demonstrated by 
anthropological materials recovered from burials at Tasty-Butak, Tursumbay, 
Khabarnoe, etc. (Debets 1954: 489-492; Ginzburg 1956a; Durnovo 1970; 
Alekseev 1964b: 22, 23; 1967; Rud' 1981). At the same time a more massive 
type of Europoid (the so-called Pamir-Fergana or Andronovo type), predom- 
inated across most of Andronovo territory in the Urals, across all of Kazakhstan, 
in Fergana, and as far as the Sayan-Altay and Minusinsk Basin (Debets 1948; 
Levin 1954; Ginzburg 1957; 1962a; Alekseev 1961a, 1961c; 1967; Trofimova 
1962a; Gokhman 1973; 1980; Rykushina 1976; 1979). In Siberia this type 
appears as a result of migration from the west, from Kazakhstan (Debets 1948: 
70, 76; Dremov 1973; 1997). On the northern periphery of Andronovo territory 
one encounters Europoid skulls with a slight admixture of Mongoloid features, 
probably due to the participation of Siberian peoples (Dremov 1972; 1997; 
Shevchenko 1976; 1980). 

The Saka of Kazhakstan and Central Asia were formed on the basis of the 
Andronovo type (Ginzburg 1951; Rychkov 1964; Alekseev and Gokhman 1984: 
21, 27, 35; Ismagulov 1963; 1970; Tot and Firshtein 1970; Gokhman 1973; 
1980). Thus, anthropological data confirm the conclusion that the Iranian Saka 
and Sauromatians succeeded the Andronovo, which was also demonstrated by 
the retrospective method. Anthropological materials also show the closeness and 
interrelationship between the Timber-grave and Andronovo populations in the 
large contact zone and the mixed character of the bearers of the Tazabagyab 
culture which has been established from the cemetery of Kokcha 3 (Trofimova 
1957; 1959; 1961; 1962) as well as the succession of the Saka population of the 
Aral Sea region with tribes of the Bronze Age (Itina and Yablonsky 1997: 
80,81). 

As far as the anthropological type of the Timber-grave culture is concerned 
new material confirms the direct genetic succession between the Timber-grave 
and Poltavka populations of the Lower Volga and shows that west European ties 
were decisive in the formation of the Catacomb and especially the Abashevo 
cultures (Shevchenko 1989: 129-130). This conclusion is of great significance 
for resolving the problem of the formation of the Sintashta and Potapovka 
populations of the 17th— 16th centuries BC. Skulls from Potapovka burials belong 
to the massive proto-Europoid type and are similar to the earlier Catacomb and 



1 70 CHAPTER ELEVEN 

genetically follow the Timber-grave and west Andronovo, but differ from 
Abashevo (Yablonsky and Khokhlov 1994: 189). 

In the 2nd millennium BC a population different in origin coexisted in 
Central Asia. In the south, in the foothills of Turkmenia, in Margiana and Bactria 
(Sapalli), as well as in Fergana (Chust, Dal'verzin) was a land dominated by 
agriculturalists; according to V. P. Alekseev (1990) this was a territory 
genetically connected with the Near East and Iran (Capprieri 1973) and also 
partially with north-west Hindustan, as in the steppes of Central Asia there was a 
merging of different populations. Skulls from the Andronovo cemetery at 
Muminabad are assigned to the east Mediterranean type (Khodzhaiov 1977: 9) 
and are close to Zaman-Baba, the steppe component of the Chust and Dal'verzin 
populations (Ginzburg and Trofimova 1972: 77). A skull from the late Andro- 
novo cemetery of Vuadil' is analogous to material from Kokcha (Trofimova 
1960: 114; 1964: 11), i.e., already mixed; another skull resembles Afanas'evo 
ones (Ginzburg 1956b). 

Other materials provide unquestionable evidence that steppe populations 
advanced to the south of Central Asia: skulls from mixed Timber-grave and 
Andronovo cemeteries at Patma-say and Karalemata-say in Turkmenia and from 
a burial on the settlement of Takhirbay 3 are proto-Europoid with Timber-grave 
and Andronovo features (Ginzburg 1959a: 105-206). A. Isakov made a signific- 
ant discovery in the cemetery of Dashti-Kozi whose materials combine elements 
of the Andronovo (Fedorovo type), and Mollali cultures: of twelve skulls, three 
are male and belong to the Andronovo type (Isakov and Potemkina 1988). This 
confirms the conclusion about contacts between aboriginal farmers and northern 
pastoralists. 

The further advance of the steppe -tribes to the south-east is demonstrated by 
the cemetery at Qawrighul/Gumugou on the shores of Lopnur in Xinjiang, where 
skulls are close to Andronovo and Afanas'evo (Han Kangzin 1994; 1998). 

Cranial analysis by B. Hemphill and A. Christensen (1994) from the Bactrian 
cemeteries of Sapalli and Jarkutan have shown that the Bactria-Margiana 
Archaeological Complex (BMAC) was formed from a population migrating from 
north-west Iran, and its creators did not migrate afterwards into the Indian 
subcontinent. This excludes the hypothesis of A. Parpola (1988) who suggested 
that the creators of the BMAC were Indo-Aryans who migrated to India. They 
also showed that the BMAC began to interact with the northern steppe tribes thus 
supporting my hypothesis. 

A migration from the north of groups of pastoralists is confirmed by data 
from Shortughai in Afghanistan. Here were found burials of the Bishkent culture, 
whose skulls, according to L. Bushe, belong to the same type as those found in 
the early cemetery of Tulkhar in northern Bactria and which have no analogies in 
the south (Francfort 1989: 211-223). T P. Kiyatkina initially suggested that the 
Tulkhar population belonged to the broad- faced Europoid type of the steppe zone 
of Eurasia; however, later she rejected this, stressing a unique Tulkhar series for 
the 2nd millennium BC (Kiyatkina 1976: 16, 17, 61). 

The progress of pastoralists to the south, into north-western Hindustan, can 
be traced in the cemeteries of Gomal and Swat, whose culture reflects the 
interaction of Bishkent and Andronovo components (Mandel'shtam 1968; 
Litvinsky 1981; P'yankova 1982b; Kuz'mina 1972a, b; 1975). Skulls from the 



VERIFICATION OF THE HYPOTHESIS 1 7 1 

Timargarha cemetery have no analogies in the Indo-Near Eastern region; rather it 
is the northern, robust, massive type close to what is found in Tadzhikistan in the 
2nd millennium BC and the Pamir Saka (Bernhard 1967: 376). T. P. Kiyatkina 
(1976: 23-25) notes the closeness of the Tigrovaya Balka skulls to Timargarha, 
bringing them together with the Timber-grave and Tazabagyab ones, but also 
stressing their difference (Tadzhik skulls have narrower faces). D. Alciati (1967) 
believes that the population that left the later cemetery of Butkara 2 of the Swat 
culture were migrants from Iran or from the area between the Caspian and the 
Aral seas. 

It is important that the modern inhabitants of the Hindukush — the Nuristani 
and Dards — preserve archaic Indo-Iranian features in language and culture 
(Fussman 1977) and often have light hair and eyes (Herrlich 1937) which proves 
the genetic succession of these populations (contrary to V. P. Alekseev (1990: 
203) who believed that the narrow faces of the Nuristani could be explained by 
temporal variability). 

Thus, the existence of two zones (Eurasian steppe pastoralists and south 
Central Asian farmers), obtained by aligning an Indo-Iranian culture with a 
definite ECT, is supported by anthropological evidence: in the south there was 
the Mediterranean type and in the north there was the Andronovo variant of the 
proto-Europoid from which descend both the nomadic and farming Iranian 
population of north Central Asia and Kazakhstan of the 8th-4th centuries BC 
(Khodzhajov 1977: 13; 1983: 100-102; Alekseev et al. 1986: 125-130). The pro- 
position of a suite of morphological forms, advanced by anthropologists, that 
finds a northern origin for the Andronovo and Tazabagyab series based on ethno- 
genetic factors assigned to a steppe ECT (Alekseev et al. 1986: 127) are in 
complete accord with our independent conclusions. 

As for Hindustan and Iranian territories, anthropologists note the mixed 
character of their modern population (Alekseev 1964a; 1981; 1990) that can be 
compared with an advance of pastoral tribes to the south. In Iran a population 
movement from east to west, from Hissar III to Hasanlu, is traced anthro- 
pologically in the Early Iron Age I, connected with the appearance of Grey Ware 
that supports E. A. Grantovsky's (1970; 1998) hypothesis about Iranian 
migration. V. P. Alekseev (1981: 199-207) showed that the majority of modern 
peoples speaking Indo-European languages belong to variants of the Europoid 
type, which developed in European territory from the Neolithic and goes back to 
a considerable degree to the Cromagnon type. Two groups form an exception: 
the Armenians, belonging to the Armenoid type, formed in the Near East, and 
part of the Indo-Iranians (groups of the modern population of Iran and north 
India) with a greater mixture of racial types, but on the whole belonging to the 
Europoid dolichocranial type that was formed in the Near East and was 
established on Iranian territory and in south Central Asia at least from the 
Eneolithic. At the same time another part of the Indo-Iranians — the modern 
northern Tadzhiks and descendants of ancient Iranian such as the Uzbeks, as well 
as the ancient Iranian population of the greater part of Central Asia and the 
Eurasian steppes (Saka), belong to another anthropological type that is charac- 
terized by a massive facial skeleton and that is related to other groups of Indo- 
European populations of Europe where this type was formed (Alekseev 1981; 
Alekseev and Gokhman 1984). Thus, Iranian-speaking peoples were formed in 



1 72 CHAPTER ELEVEN 

an anthropological environment composed of two Europoid components that 
differ in their genesis. Moreover, paleoanthropological evidence provides "real 
proofs of the ancient formation of northern groups of Iranians on the basis of 
massive Europoid combinations", and "the natural habitat of the northern, more 
massive component included the vast territory of the south Russian steppes, 
Kazakhstan, and northern areas of Central Asia, including north and south 
Tadzhikistan, that formed a far southern periphery" (Alekseev 1981: 203-204). 
The main component of the formation of the Saka was the Andronovo type. 
From this it follows that the Eurasian steppe nomadic Saka were not immigrants 
from the Near East but direct descendents of Andronovans, and the mixed 
character of the Indo-Iranian-speaking populations of Iran and India is the result 
of a new population spreading among aboriginals with whom a new language is 
probably to be associated. 

This conclusion is confirmed by the evidence of Indo-Iranian tradition. The 
Aryans in the Avesta are tall, light-skinned people with light hair; their women 
were light-eyed, with long, light tresses (Yasna 26; Yast 5.7, 15, 64, 78, 126; 
13.107; 17.11) (Dhalla 1922: 23). 

In the Rigveda light skin alongside language is the main feature of the 
Aryans, differentiating them from the aboriginal Dasa-Dasyu population who 
were a dark-skinned, small people speaking another language and who did not 
believe in the Vedic gods (Elizarenkova 1972: 11; 1998: 433; Parpola 1988). 
Skin color was the basis of social division of the Vedic Aryans; their society was 
divided into social groups varna, literally 'color'. The varnas of Aryan priests 
(brahmana) and warriors {ksatriya or rajanya) were opposed to the varnas of the 
aboriginal Dasa, called 'black-skinned' {Rigveda 1.130.8). Modern researchers 
propose that the Dasas were Dravidians, creators of the Harappan culture and 
script (Bongard-Levin and Gurov 1988: 65; 1990; Parpola 1988). Their home- 
land in the Punjab is proved by the Brahui who have preserved Dravidian lan- 
guage in the north and a cultural vocabulary reflecting an ancient acquaintance 
with the productive economy and developed crafts; moreover, some terms were 
later borrowed into the Indian languages as the newcomers adapted to a new 
ecological niche (Sankalia 1973; Kuiper 1991). Proto-Dravidians are genetically 
connected with the Near East, and are probably related to the Elamites and are 
assigned to the Europoid Near Eastern anthropological type (Alekseev 1990: 
170). The Harappan population also had an Austronesian admixture. The greater 
part of the modern Indian population belongs to the Indo -Afghan type, forming a 
branch of the Mediterranian group. Genetically they go back to the Mohenjo- 
Daro, Mundigak and Sialk cultures (Mendez 1966). The representatives of the 
Munda people could also be included among the Dasa; they are assigned to the 
Austronesian language family, distinct from the south of Hindustan. Anthropo- 
logically, the Munda belong to the Australoid-Negroid population and are 
characterized by dark-skin pigmentation and a flattened face. 

Indian culture formed as the result of the active interaction of three aboriginal 
ethno-linguistic traditions: Proto-Dravidians of the Punjab, Munda of the south 
and Tibetan-Burmese of the north-east and the immigrant Indo-Aryans (Voro- 
b'ev-Desyatovsky 1956; Bongard-Levin 1979; 1981; Bongard-Levin and Gurov 
1988; 1990). In modern India the north-western population is much lighter than 
the southern; moreover, representatives of lower castes have darker skin color, 



VERIFICATION OF THE HYPOTHESIS 1 73 

while higher ethnic-caste groups, who trace their genealogy from the ancient 
Aryans, are lighter, and among the Brahmans even light-eyed people are 
occasionally found. 

V. P. Alekseev (1990: 161, 206) considers "the thesis of T. V. Gamkrelidze 
and V. V. Ivanov about Iranians separating through western areas of Central 
Asia and their movement around the Caspian Sea to be insufficiently 
demonstrated." He localizes the Indo-Iranian homeland in the north and suggests 
that they advanced from Central Asia. 



Toponymic evidence for the Indo-Iranian homeland 

The geographic zone of the ancient seat of the Indo-Iranians is indicated by the 
absence in the Indo-Iranian languages of the common Indo-European words for 
'spruce' and 'bog'. This makes it possible to localize them in the steppe 
(Schrader 1913). The suggested identification of the legendary river Ra (Sanskrit 
Rasa, Iranian Rarjhd) with the Volga and the Ripa mountains with the Urals 
(Marquart 1938; Grantovsky and Bongard-Levin 1970; 1998; Grantovsky 1976 
supported by A. I. Dovatur, D. P. Kallistova et al. (1982: 248-249) and N. L. 
Chlenova (1983: 56-60; 1989)) is extremely important for locating the Indo- 
Iranian homeland. V. Miller (1887) recognized the Iranian etymology of a 
number of geographical names in the North Pontic, later augmented by V. I. 
Abaev (1949). A large number of Iranian hydronyms with the root don (Iranian 
'water') has been revealed for the drainages of the Dnieper, Desna, Seim, 
Severny Donets and the Poltava region (Toporov and Trubachev 1962; Strizhak 
1965). The Indo-Iranian treatment of toponyms in the North Pontic has also been 
supported in a series of works by O. N. Trubachev (1975; 1976; 1999) and L. A. 
Lelekov (1980) while E. A. Grantovsky and D. S. Raevsky (1984: 47-62) agreed 
that the Iranian element is "the only undisputed (element) in the North Pontic." 
S. S. Berezanskaya (1982: 206-209) in charting the hydronyms of the Upper and 
Middle Dnieper area on an archaeological map showed that the area of the 
Iranian names does not coincide with the distribution of the Scythian culture but 
correlates completely with the territory of the Timber-grave culture, which is a 
strong argument in favor of assigning a Proto-Iranian identity to the Timber- 
grave people. 

Indo-Iranian toponyms are also found on the Middle Volga and the Urals 
(Popov and Loyfman 1962). N. L. Chlenova (1983a; 1984) plotted the toponyms 
and demonstrated that Iranian toponyms were spread over Timber-grave and 
Andronovo territories, but part of the Andronovo toponyms can only be 
interpreted as Indo-Aryan; here also was included the Altai by A. M. Maloletko 
(1986: 70-75). These data are extremely important for the final resolution of the 
problem of the cultural and ethnic attribution of the Fedorovo complex. Due to 
the fact that Indo-Iranian toponyms of the pre-Scythian period have been found 
on the territory populated only by Fedorovo tribes, the hypothesis identifying the 
Fedorovo population as Ugrian that has been proposed by V. N. Chernetsov must 
be rejected, and the hypothesis of the Indo-Aryan attribution of Andronovans can 
be supported. 



1 74 CHAPTER ELEVEN 

The same conclusion can be drawn from the toponymic analysis of the 
mountain regions of the Pamirs and the Hindukush where Andronovo-Fedorovo 
burials and petroglyphs have been recently discovered, including chariot 
paintings that are sometimes found on mountain passes; these probably indicate 
the direction of Indo-Aryan movement to the south. G. Morgenstierne (1929- 
1938; 1973) studied relict Indo-Iranian languages here and outlined a large layer 
of ancient toponyms (Edelman 1968). The Iranian toponym dara- 'river' is 
distributed from the Tian-Shan and Pamirs to the south of Tomsk and 
Novosibirsk and further to the borders of Mongolia, Pakistan and Afghanistan 
(Dulzon 1964; Pospelov 1980: 122). A substratum layer is present in the 
toponyms of the Pamirs which include Iranian names (Pakhalina 1976: 178-182; 
Steblin-Kamensky 1976: 182-185; Rozenfeld 1980: 157-162; Grunberg 1980: 
165-169; Bailey 1958). E. Thomas (1883: 357-386), P. Bhargawa (1971), T 
Elizarenkova (1972: 12-13; 1982: 4; 1989: 433, 440-443) and M. Witzel (1999a; 
b; 2001) analyzed the toponyms of the Rigveda that were applied to eastern 
Afghanistan and North-West India thus demonstrating Indo-Aryan migration 
from Central Asia through the mountain passes of northern Afghanistan and 
Baluchistan to the north-west of the Indian subcontinent. 

The toponyms of the Hindukush and the river -names of north-west India 
possess Indo-Aryan etymologies and they are often reflected in the Vedas 
(Elizarenkova 1989: 438). However, in this region of Aryan toponyms, the 
names of the aboriginal Dravidian population are also encountered. This 
language is spoken by the Brahui, a relict of the aboriginal population of India 
(Sankalia 1973). 



Indo-Iranian traditions on the homeland 

The mythology of both the Iranian and Indian peoples preserve tales about their 
arrival from a legendary northern homeland, which have been analyzed by G. M. 
Bongard-Levin and E. A. Grantovsky (1976; 1983). 

In Iranian texts, the idea about the kinship of all Iranian-speaking languages 
is reflected in a legend of how the ancestor of the Iranians divided the land 
between three sons: Sairima, the forefather of Sauromatians (who dwelt in the 
historic period from the Don to the Urals), Tur, from whom the Turians 
originated (the northern part of Central Asia was called Turan), and the younger 
son Iraj, the ancestor of the Iranian population (Christensen 1934). In the Avesta 
there is also a tale about the migration of his people to the south led by Yima; 
this migration was caused by overpopulation and the fall of temperature in the 
homeland (Videvdat 2.3). 

A mythologeme of the Avesta also points to a northern homeland: in the 
ArdvTsur Yast (V) the goddess Anahita is depicted in a beaver's coat. Such 
attribution of a deity could originate only as a result of more ancient ideas about 
the association of the goddess, who is protectress of waters, with this animal that 
inhabits water (Rapoport 1971: 31). Beaver bones are widely represented in the 
materials of Andronovo settlements in the Urals (Kipel', Novo-Burino), on the 
Tobol (Alekseevka), in central (Shortandy-Bulak) and eastern Kazakhstan (Usf- 
Narym, Malokrasnoyarka, Trushnikovo), on the Ob — in the Pichugino cemetery 



VERIFICATION OF THE HYPOTHESIS 1 75 

there was found the skull of a beaver (Krivtsova-Grakova 1948: 102; Chernikov 
1960: 60; Tsalkin 1972: table 6; Makarova 1974: 203; 1978: 138; Smirnov 1975: 
32; Martynov 1964a: 240; Potemkina 1985: 309). The beaver's natural habitat in 
the Old World was limited to the forest zone; discoveries in Kazakhstan fix the 
southernmost borders of the distribution of this species; at present the beaver is 
not met in this area (Skalon 1951; Afanas'eva 1960; Kozhamkulova 1969: 23; 
Lychev 1977; Lavrov 1981, etc). This means that the mythological image of the 
water goddess could have been created only in this zone. It is necessary to 
mention that the cult of the mother -beaver was also developed among Finno- 
Ugrian peoples, neighbors of the ancient Iranians. The connection of the mother- 
goddess with the inhabitant of water is found among other steppe peoples: the 
Scythians had a snake -legged nymph, a daughter of Borisphen (the Dnieper), the 
Narts' prime ancestress was Dzerassa, the daughter of Donbatyr, who lives in the 
water as a water turtle (Kuz'mina 1977: 21). A common Indo-European word 
derived from the lexeme 'water' denotes in the Indo-Aryan language a cult water 
animal in general. These data, in spite of the opinion of T. V. Gamkrelidze and 
V. V. Ivanov (1984: 529-531), make it possible to suggest the ancient origin of 
this cult and indicate its formation in the Eurasian forest-steppe zone close to the 
Finno-Ugrian peoples. 



Art and mythology 

The analysis of art is one of the means of verifying the Indo-Iranian identity of 
the Andronovo culture. Geometric ornament predominated on the ceramics of 
the pastoral tribes of the Eurasian steppes in the Bronze Age. This has led to the 
erroneous conclusion that their art was aniconic. However, samples of the art of 
people living in the region are quite numerous, and are presented in 
anthropomorphic and zoomorphic images on monumental sculpture, ceramics, 
small plastic art and petroglyphs. 

Monumental sculpture includes large stelae attributed to the Andronovo 
cemeteries and mines of central and eastern Kazakhstan that depict schematic 
figures of rams, horses, Bactrian camels and bears with head raised towards the 
sky (Margulan 1979: 283, 285, 291, 292). A. H. Margulan notes that they are 
typologically similar to artwork in stone. 

The representations on Andronovo and Timber-grave ceramics have been 
discussed in Chapter 8 (on transport). In addition to chariots there are single 
anthropomorphic figures such as a skier on a vessel from the Voronezh region 
(Vinnikov and Sinyuk 1990: fig. 34). 

Small objects of plastic art may be made of stone, clay and bronze and 
include anthropomorphic and zoomorphic images. 

Anthropomorphic images are found on stone scepters or pestles with mask- 
like heads (Fig. 55). A scepter from the Semipalatinsk region (Hermitage 
Museum; No. 1648/1; Margulan 1979: 292, fig. 222) that is oval -square in 
section has a protruding area on top below which was a mask; a scepter from the 
river Tuy, a right tributary of the Irtysh (Tobol'sk museum) comprised a round 
rod with a roller on one end terminating with the protruding head of a man 
(Moshinskaya 1952: 46-48, fig. 17 a; 1976: 56, table 6); from the river Ir, a right 



1 76 CHAPTER ELEVEN 

tributary of the Ishim (Omsk museum) comes a long round rod with a roller and 
is flattened on top depicting two weakly protruding heads, similar in form, and 
joined at the back of their heads (Moshinskaya 1952: 48, fig. 176.1); a headed 
pestle is known from the Nura river of the Tselinograd region (Petropavlovsk 
museum) (Zdanovich and Moshinskaya 1973: 199-202, fig. 1; Moshinskaya 
1976: 55, table 5); on a precipice overlooking the Ul-Zhilanshik river of the 
Turgay region (Turgay museum), a small stone smoothing-iron has been 
recovered with a protruding face on the butt-end (AO 1974: 492, fig. 491). A. 
Kh. Margulan (1979: 307, fig. 224, table 22. 2) was probably describing the 
same object although he claimed it was found on a settlement on Lake Akkol on 
the river Dzhilanchik and it is in Alma-Ata Art museum of Kazakhstan. In 
addition, two analogous naked male figures in sitting postures with raised heads 
are known: one from the Tobol river near Kustanay from the Kustanay museum 
(Zdanovich and Pleshakov 1981: 258-261, fig. 1.2) and the other from the 
Atbasarka river of the Akmola region, now in the Eastern Department of the 
Hermitage (Shkoda 1992: 58). A number of new discoveries have been made in 
Kazakhstan which have not yet been published. Finally a naked male figure of a 
skier on a bronze knife from grave 2 at Rostovka in the Irtysh area belongs 
stylistically to this group (Matyushchenko 1970: 103, fig. 33). 

The objects described are technologically and stylistically uniform. They 
were made by pecking out brown sandstone or green serpentine. The face was 
modeled on several protruding planes. It is possible to state that its prototypes 
were carved wooden sculptures. The relief stresses superciliary arches, 
protruding skulls and a heavy chin, wide straight protruding nose, low squinting 
forehead, deepened cavities for eyes (only on the head from Tuy did they drill 
round cavities), the lips expressed in relief and the mouth comprise a straight 
line, that is absent in the Nura figure; the heads from Tuy and Ir lack the lips in 
relief, and figures with ears marked with rounded projections. The facial type is 
Europoid and close in all indices to the Andronovo (some Mongoloid admixtures 
are possibly found only on the skier from Rostovka). A round helmet or hat is 
found on every head but the Semipalatinsk one; it is slightly marked on the 
Akkol figure; the Nura and Rostovka show a cut from behind. 

Initially V. I. Moshinskaya (1952: 50) dated the anthropomorphic figure 
found near the Irtysh to the 1st millennium BC. L. R. Kyzlasov (1956) assigned 
it to the Andronovo culture on the basis of its similarity to pestles with 
anthropomorphic figures which are now attributed to the Okunevo culture on the 
basis of closed finds. In 1976 V. I. Moshinskaya suggested the dating of this 
group to the second half of the 2nd millennium BC. However, the age of the 
sitting figure from Kustanay was lowered to the end of the 3rd / beginning of the 
2nd millennium BC (Zdanovich and Pleshakov 1981: 261) and the analogous 
one from Atbasarka was set even to the middle of the 4th / middle of the 3rd 
millennium BC (Shkoda 1992: 58). But the stylistic unity of the whole series 
permits us to consider its creation within a single cultural sphere in a relatively 
short period of time. Decisive for cultural attribution are the small smoothing- 
iron with a mask and flat adze of Andronovo type on the Andronovo settlement 
of Akkol and a knife from Rostovka (which combines the images of a man and a 
horse; Fig. 68: 3) that permits us to make comparisons of the zoomorphic plastic 
art. It also allows us to consider the burial complex of the Chinese empress Fu 



VERIFICATION OF THE HYPOTHESIS 1 77 

Hao in the imperial cemetery at Anyang in Xiaotun ("Henan chutu" 1981: 147, 
148, fig. 136-181). This contained vessels with the empress's name, stone figures 
of Europoid men in round caps squatting or sitting on the knees with raised head 
treated (as can be seen from the reproduction) in a manner close to the Irtysh 
series (Muller-Karpe 1980: 366, fig. 123-126, 137, 145). The assemblage also 
includes a curved-back knife with the figures of horses (Muller-Karpe 1980: fig. 
56, 140, 141) with correspondences in the zoomorphic plastic arts of the steppe 
and forest-steppe; nephrite rings and plates may be compared with those from 
Seyma, and a round bronze mirror is of Andronovo type. The burial dates to the 
13th century BC. 

Images of the horse occupy first place among the zoomorphic steppe figures 
in the plastic arts (Fig. 68). A stone horse-headed scepter was found near Malaya 
Cheka in the district of Chelyabinsk (Drevnosti 1991: 54, fig. 32). Horse heads 
are on a bronze spear shaft from Dzhetygora in northern Kazakhstan 
(Maznichenko 1986: 152-154), on stone scepters from eastern Kazakhstan in the 
Semipalatinsk district (Chernikov 1960: 24, fig. 21), near Bukhtarminskaya 
station (Slavnin 1949), and from the settlement of Volchiy in the Omsk region 
on the Irtysh (Moshinskaya 1976: 70, fig. 11). A pair of opposed horses forms 
the upper part of the handle of a bronze curved-back knife from Dzhumba from 
the Kalbin ridge in eastern Kazakhstan (Chernikov 1960: tables 1, 14.5); small 
horses are depicted on the ornamented handles of two daggers from the Karakol 
II hoard (Fig. 68: 5,6) in Semirech'e (Vinnik and Kuz'mina 1981: fig. 1); a pair 
of horses decorates a round temple ring with gold facing from the Andronovo 
cemetery of Mynchunkur (Fig. 68: 4) in Semirech'e (Akishev K. and Akishev A. 
1983: fig. 48). Thus objects with horse images are found mainly in eastern 
Kazakhstan, Semirech'e and the Irtysh. 

Similar objects are known northwards down the Irtysh in the Rostovka 
cemetery on the handle of curved-back knife with a horse depicted pulling a 
skier (Matyushchenko 1970: 103-105, fig. 33, 34) and eastwards in the Altai on a 
horse-shaped stone pendant from the Krotovo culture settlement of Saranin 2 in 
the Omsk region (Glushkov 1986: 196). Finally, analogous to the pair from 
Mynchunkur are a horse and mare on the butt of a Seyma knife (Bader 1970: fig. 
52). Horses are depicted in the same way: they exhibit a large raised head with a 
thick mane and protruding fringe, drooping belly, thick short legs and fluffy long 
tail. Their species has been defined as the wild Przhevalsky horse (Bader 1970: 
104; Kiryushin 1987: 1 17) but the horses are domestic, as there is a bridle on the 
Seyma mare and a horse from Rostovka. Their appearance suggests that they 
belong to an ancient Kirgiz breed now found in Semirech'e, eastern Kazakhstan 
and they are ancestors of the modern horses of the Mongol breed (Sokolov 
1959). In the past they were distributed from the Asian steppes to China and 
Japan. 

Kulans-dziggetais, now found in Xinjiang and Central Asia and in ancient 
times inhabiting the whole Eurasian steppe from the Ukraine to Mongolia and 
also the Trans Caucasus and the Near East, or tarpans (Equus Gmelini), wide- 
spread across the steppes, found from the Ukraine to Central Asia till the 1 8th 
century (Pallas 1786: 82, 83), are depicted on two single -bladed knives from the 
Altai (from the Elunino cemetery of the Elunino culture on the Ob; Kiryushin 
1987: 101, 117, fig. 2, 4) and from Ust'-Mut (Pyatkin and Miklashevich 1990, 



1 78 CHAPTER ELEVEN 

fig. 1, 2). Both animals differ from the typical horse: they are more massive, 
have large foreheads, a short upright mane without a fringe and with long 
pointed ears (Sokolov 1959). These features were emphasized by the Altai 
craftsmen. 

The second most frequent image in plastic art is the wild stone ram (argali) 
and mountain goat. The natural habitat of these animals includes the eastern 
districts of Central Asia, Kazakhstan and the Altai. Rams decorate daggers from 
the Karakol II hoard (Vinnik and Kuz'mina 1981, fig. 1) and a single -bladed 
knife from Turbino (Bader 1964; 115-123, fig. 113). A mountain ram is placed 
on the celt from Rostovka in the Irtysh (Matyushchenko and Lozhnikova 1969: 
table 6). 

Fragments with the expressive image of Bactrian camels that inhabited all 
Kazakhstan and Central Asia are found from the Irtysh (Moshinskaya 1976: 70, 
72, fig. 12), from the Kusak river of the Djezkazgan region (Margulan 1979: 
232, fig. 214) and from the Andronovo settlement of the early Kozhumberdy 
type at Ushkatta 2 in the Urals (Kuz'mina 1962b: fig. 3, 4; 1963a: 38-46, fig. 55. 
8, 90). 

Stylized bird figures terminate a spear shaft with horses from Dzhetygyr in 
northern Kazakhstan (Maznichenko 1986: 152-154). A beast, probably a tiger or 
snow leopard, inhabitants of Semirech'e, eastern Kazakhstan and the Altai, are 
found on a spear shaft reinforced with a forked rod found on the Irtysh south of 
Omsk (Chernykh and Kuz'minykh 1989: 67, fig. 31.1). Another schematic stone 
image of an undefined animal is held in the Alma-Ata museum (Moshinskaya 
1976: 70). 

Hence an analysis of the plastic art originating from the Andronovo culture 
indicates: 

1. Some images are on knives (Dzhumba, Elunino, Ust'-Mut, Rostovka, 
Turbino, Seyma); they belong to one type: curved -back massive cast objects with 
protruding sharpened blade on the end, thickened back and ornamented handle. 

2. In spite of differences in material (bronze, stone and clay), the stylistic 
manner is unique with the images of horses most strictly canonized. 

3. The absolute majority of objects originate from eastern Kazakhstan, 
Semirech'e and the Altai. 

4. The repertoire of the craftsmen is limited to several species of local fauna: 
the domestic horse of the Kirgiz breed, kulan, Bactrian camel, mountain ram, 
mountain goat and some wild feline (snow leopard or tiger). 

5. The camel's head from the Omsk museum is made of quartz sandstone 
originating from north-eastern Kazakhstan (Moshinskaya 1976: 72). 

This leads to the conclusion that the appearance of such images outside the 
borders of the limited area proves them to be imports from the Semirech'e- 
eastern Kazakhstan region and indicates a Seyma-Turbino origin. 

O. N. Bader (1971: 98-103) remarked that the animal species on Seyma and 
Turbino knives represent steppe and mountain fauna of Central Asia. V. I. 
Moshinskaya (1976: 69-79, 109) analyzed Ural and west Siberian sculptures, 
dated a number of samples to the Bronze Age, and supported the idea that Seyma 
objects were produced in the Irtysh area after she compared the latter with plastic 
art from the settlement of Samus' IV (Moshinskaya 1976: 58-60) with Seyma 
type metal. A. A. Formozov (1976: 5-7) stressed the difference of the prevailing 



VERIFICATION OF THE HYPOTHESIS 1 79 

zoomorphic images among the forest hunters and the steppe pastoralists, 
comparing the latter with the decorative art of the Okunevo culture and saw its 
influence at Samus' and Rostovka. The depiction of a mountain goat on the celt 
from Rostovka made V. I, Matyushchenko (Matyushchenko and Lozhnikov 
1969) question the southern, that is Altaic, origin of Rostovka metallurgy. That 
was supported by M. F. Kosarev (1974: 115). The Karakol II hoard complex 
forced D. F. Vinnik and myself (1981) to reject the prevailing assumption about 
a connection between the single-bladed curved-back knives and the Karasuk 
culture (Artamonov 1973: 86), and show their closeness to Semirech'e and 
eastern Kazakhstan prototypes, suggesting a Seyma-Turbino production center in 
this region. 

On the basis of other evidence (metal analysis) E. N. Chernykh (1970: 155- 
173) concluded that there was an eastern origin for ore Group 1 of the Seyma 
bronzes, which was demonstrated by his further research with S. V. Kuz'minykh 
(1989: 270). The authors see two sources of this trans-cultural phenomenon: the 
Altai (Elunino, Kanay, Ust'-Bukon) steppe metallurgists and horsemen and the 
forest hunters who engraved in stone and bone in the east Siberian taiga 
(Glazkovo). 

B. N. Pyatkin and E. A. Miklashevich (1990: 150-152) followed A. A. 
Formozov in assigning to the Okunevo culture the whole series of zoo- and 
anthropomorphic images. Their comparison of the images of the horse in its 
special stylistic manner in the plastic art and petroglyphs of Semirech'e, eastern 
and partly central Kazakhstan, seems rather convincing but the attribution to the 
Okunevo culture is not sufficiently demonstrated. 

What is the origin of this art style? From what we have seen so far it follows 
that most material originates from Semirech'e and eastern Kazakhstan, the center 
of tin mining for bronze production, which provided an enormous area for 
mining. Prototypes of curved-back knives and spears with forked-reinforced 
shaft specific for Seyma-Turbino metallurgy were found in the same district. 
They can be derived from an archaic spear from the Semipalatinsk dunes with a 
leaf-shaped blade without a midrib and three ridges for enhancing the firmness 
of a curved socket (Chernikov 1960: table 65. 1). 

There is no need to trace Seyma stone and bone arrows from Trans Baikal: 
they are widely distributed across the whole Andronovo territory. As for neph- 
rite, its deposits are in the Tian-Shan, Pamirs and in eastern Turkestan. Lapis 
lazuli was extracted from the same mines. A southern rather than an eastern 
origin of Seyma stone ornaments is probably seen by the discovery of a bead of 
light-green stone in grave 9 in the Kanay cemetery in eastern Kazakhstan, and 
cylindrical segmented paste beads in graves 3 and 7 (Chernikov 1960: 32, 34, 
tables 19.3,4, 20b), beads of green and white nephrite from Rostovka, cross- 
shaped turquoise beads in the burial of Sopka 2. Turquoise beads are found in 
many Petrovka burials; in typology they are analogous to those of Central Asia 
during NamazgaV and VI (Kuz'mina 1988b: 51, 52; Zdanovich 1988: 138). 

In the south of Central Asia there was a continuous tradition of depicting 
animals in plastic art in 4th to 2nd millennia BC. Metal pins with ox, cow with 
calf, goat, mountain ram, bird heads or figures as well as compositions with 
humans are especially noted from southern Turkmenia and Bactria (Kuz'mina 
1966: 80-82, table 16; Sarianidi 1977: 83, 85, fig. 43, 44; Askarov 1977: tables 



1 80 CHAPTER ELEVEN 

1, 6, 5, 6). The most north-easterly discovery of such a type comes from the Hak 
hoard in Fergana, immediately neighboring Semirech'e. V. G. Shkoda's (1992) 
opinion on the stylistic similarity between a figure from the Atbasarka river and 
the anthropomorphic image on a stone vessel from Sarazm near Samarkand is 
important. They, as well as the pin with horse on the top from Zardcha-Halifa 
(Fig. 68: 1), demonstrate contacts between pastoralists and southern farmers. 

Thus, it is possible to imagine that the origin of plastic art in the steppes was 
stimulated by southern contacts with bronze smiths of Semirech'e and eastern 
Kazakhstan, who delivered tin or bronze to the farmers in exchange. This is 
shown by the appearance of tin alloys in Turkmenia and in Bactria in particular 
in the 2nd millennium BC along with traditional arsenic alloys (Kuz'mina 1966: 
91; Askarov 1977: 123, 124) and ready made objects of the Andronovo type 
(Fig. 53). 

What is the cultural attribution of the creators of this art and the whole 
Seyma-Turbino complex? S. S. Chernikov (1960: fig. 21, table 14, 5) related the 
material from eastern Kazakhstan to the Andronovo culture. Present day 
arguments in favor of this attribution are: 

1. The absence of Okunevo or sites of some other culture; only Andronovo of 
the Fedorovo type is found in this territory. 

2. There is a probable genetic connection between the Fedorovo population 
of eastern Kazakhstan and the previous (non Okunevo) population represented in 
the destroyed burial of Ust'-Bukon' (where one finds a cist and pots with 
rounded shoulder and archaic horizontal herring-bone pattern and impressions 
(Chernikov 1960: 16, 17, table 6.1, 2)) and the Kanay cemetery that displays the 
typical combination of archaic weakly profiled vessels with zonal ornament 
(comparable to Ust'-Bukon' and to a certain extent to Samus') with typically 
Fedorovo pots with rounded shoulder, comb stamp decoration, including oblique 
triangles (Chernikov 1960: table 19, 3, 5). 

3. In archaic grave 9 two extended burials with raised knees and with ocher 
were found inside the enclosure. Skulls from the grave belong to the Andronovo 
type (Ginzburg 1956: 239). Fedorovo utensils were found on the settlement, only 
some fragments preserving Eneolithic traditions (Chernikov 1960: tables 17, 18). 

4. The discovery of a temple ring with two standing horses in the Mynchun- 
kur burial in Semirech'e, analogous to horses on a knife from Seyma (Fig. 68: 2), 
is of decisive importance. The cemetery belongs to the early Semirech'e sites. In 
grave construction (cist) and ceramics with rounded shoulder and saucer it can 
be compared to the Fedorovo cemeteries of eastern Kazakhstan (Karabaspakova 
1991: 14). The form of temple ring is typical of those found in Fedorovo 
cemeteries (Komarova 1962: 66; Avanesova 1975b: fig. 2; 1991). The natural 
region of cast rings wrapped with gold foil includes northern Kazakhstan 
(Borovoe) and the Ob (Prigorodnoe, Volchikha, Kytmanovo, Vakhrushevo, 
Krasny Yar and Preobrazhenka). 

It has been proposed that the invention of tin bronze, the most important 
innovation of the Bronze Age, led to the rise of the Andronovo Fedorovo 
population that inhabited the area of eastern Kazakhstan and Semirech'e which 
was rich in cassiterite deposits; it also stimulated the rapid growth in bronze 
casting here and the creation of new types of bronze objects (spears with forked- 
rod reinforcement^ single-bladed curved-back knives, cast temple rings, etc.). 



VERIFICATION OF THE HYPOTHESIS 1 8 1 

The rich cassiterite deposits led to active contacts between early Fedorovo 
tribes and their neighbors. They exported tin or bronze bars to the south. An 
original decorative art was formed in the steppes under the influence of the 
zoomorphic and anthropomorphic plastic art of southern farmers (probably 
through neighboring Fergana). 

The main flow of this production of Fedorovo metallurgists was from 
Semirech'e and eastern Kazakhstan to the north. Manufactured objects were 
sent, not raw material, which accounts for their similarity over the vast Eurasian 
territory. The main trade route was the Irtysh, which is indicated by the 
concentration along the Irtysh of a large number of finds and the appearance of 
such sites as Rostovka and Chernoozer'e. From the Irtysh over the forest and 
steppe border goods were transferred further to the west to the Urals (Turbino) 
and the Volga (Seyma). Another part of the export trade went to the north-east 
(perhaps round the Altai uplands), along the Irtysh and further through the 
Kulunda and Baraba steppes to the Altai foothills, and then to the north along the 
Ob to the taiga zone where we find the settlement of Samus' IV. 

The spread of Seyma metallurgy and art in the Altai region and pre-taiga 
Siberia is connected with the Andronovo Fedorovo tribes. This can be shown 
through several lines of evidence. On the upper Ob the Elunino culture was the 
aboriginal culture during the Early Bronze Age, with radiocarbon dates of 
1610±30 BC and 1680±75 BC. The synchronous spread of bronze objects and 
zoomorphic plastic art of the Seyma type, horse sacrifice and sporadic incidents 
of cremation (Elunino, Komsomol'skiy Mys) are known from the cemeteries of 
this culture. At Elunino and Tsygankova Sopka we find several examples of 
violent death among representatives of the local population; their anthropological 
type with its Mongoloid admixture differs from the Europoid western intruders 
(Kiryushin 1987: 100-125). 

In the Baraba steppe V. I, Molodin has described the local Krotovo culture. 
Here the spread of Seyma metal types is also connected with the Andronovo 
Fedorovo population: in Preobrazhenka III Fedorovo graves cut a Krotovo layer 
and they were covered by Irmen kurgans; at Sopka 2 Fedorovo ceramics were 
present on a Krotovo site (Molodin 1977: 64-68, tables 11, 7). Finally Fedorovo 
vessels were found in the Samus' IV settlement, that yielded Seyma type molds 
(Matyushchenko 1973a: 10, 42, fig. 6, 2). 

Acquiring horse and metal from Fedorovo people the Siberian peoples 
mastered bronze-casting and later in ore rich districts, first of all in the Altai, 
they formed their own local metallurgical centers. Horses and chariots as well as 
Andronovo-Seyma bronzes (resulting in local replicas) and Andronovo gold 
temple rings reached China through Xinjiang and northern Chinese tribes. 
Copies were then created in China. 

Several images were known in Andronovo decorative art: horse, ox, Bactrian 
camel, ram and goat. The same animals are popular on petroglyphs. The dating 
of these to the Bronze Age is based on the following: 1) stratigraphy and palimp- 
sests; 2) discovery of plates with images in graves, for instance at Tamgaly; 3) 
discovery of a cultural layer on a settlement connected with petroglyphs; 4) the 
changing of styles and themes; 5) the closeness of images on stones and dated 
material objects; 6) technique of execution; 7) weathering and staining 
(Bernshtam 1962; Kadyrbaev, Mar'yashev 1977; Medoev 1979; Sher 1980; 



1 82 CHAPTER ELEVEN 

Maksimova et al. 1983; Martynov et al. 1992; Samashev 1992; Rogozhinsky 
1991; Mar'yashev, Goryachev 1998; 1999; Novozhenov 1994, Tashbaeva 1999). 

Petroglyphs related to the Andronovo culture were found in the Balkhash 
area, in central and eastern Kazakhstan, in Central Asia, including the Pamirs. 
They are especially numerous in Semirech'e, where the two largest complexes 
are found: Tamgaly and Saymaly-Tash. Saymaly-Tash has 10,000 stones with 
drawings of different periods. It is near a mountain lake 3,400m high and 
difficult to reach. Images of horse-drawn chariots, battle scenes, ritual dances, 
and solar images form the most important group of petroglyphs of the Bronze 
Age. 

I have proposed a widely accepted hypothesis regarding the interpretation of 
Eurasian steppe art on the basis of Indo -Iranian mythology (Kuz'mina 1976c). 
The image of the horse, chariot, chariots with two riders (Asvin twins), a chariot 
with a fantastic team, Bactrian camel, ram, ritual plowing scene, heroes in battle 
(Fig. 34, 55.2, 56.3) can be explained by Indo-Iranian tradition (Kuz'mina 
1963a; 1971a; 1977; 1986b; Kadyrbaev and Mar'yashev 1977; Sher 1980, etc). 
Images of a sun-faced character on the Saymaly-Tash and Tamgaly petroglyphs 
(Fig. 55.1, 3; 56) are probably the images of the most ancient Indo-Iranian 
Mithra = Sun God. The word sun in the Parthian language originates from this 
name. Another function of the god is being a guardian of a treaty and order 
(Gershevitch 1959). "Mithra, with eyes open, observes people" {Rigveda 3.59.1, 
4); he is depicted either with one eye, i.e., the sun, or in the 'Mihr Yast' (10) in 
the Avesta he bears the epithet 'thousand-eyed'. The same image is present on a 
scene of worship and ritual dances (Fig. 56.2). A solar figure on the back on an 
ox (Fig. 56.4) probably symbolizes the cosmogenic myth concerning Mithra 
sacrificing the primal ox. The plowing scene is the common European rite of the 
first furrow, that is connected with an incantation {Atharvaveda 3.17.2, 48) and 
myth how Ahura-Mazdah handed a golden plow to the first Iranian king Yima 
(Videvdat 2.2, 16, 18). The Scythians and Saka knew the myth about the golden 
plow falling from the sky (Herodotus 4.5-7; Quintus Rufus 7.8, 18). 

The chariot, the most frequent attribute of a deity in the Indo-Iranian 
pantheon, is the most popular image in the Rigveda. For example, in Mandala V 
Agni (1, 2, 5, 10, 13), Indra (29, 30-32, 35, 43), the Asvins (41, 43, 73-75, 77), 
Maruts (53-61), Mitra-Varuna (63, 65), Usas (80) and others are in chariots. The 
epithets of the celestial chariot indicate that it is high, with solid rims, pleasant, 
beautiful, splendid, bright, golden, easy-going, fast-running, tireless, hard to be 
reached, quick, quick as thought, running like the wind, thunderous, prey- 
seeking, victory-bringing; the horses are described as fast, tireless, stately, 
beautiful, with straight back, dun, white, red, bird-horses. In the Avesta horse- 
drawn chariots are also attributes of the gods, of Anahita and Mithra in particu- 
lar. The latter's horses are "terrible, fast-running, they carry the fast chariot, 
stretching out raw-hide reins." This reflects the enormous significance of the 
chariot in Indo-Iranian culture, and recalls the image of the chariot from Thor on 
the Indus river near the pass leading from the north into India. 

Another image is the Bactrian camel which is an incarnation of the Iranian 
VsrsBragna (Behram Yast, 3) related to Indo-Aryan Indra-Vrtrahan. The ram in 
the Avesta is the incarnation of Farnah, the personification of glory and 
charisma; he is still worshiped in the Pamirs. The horse, ox, ram, and goat cult 



VERIFICATION OF THE HYPOTHESIS 183 

reflect the role of these animals in sacrifices and in mythological texts in which 
the gods are compared to these animals or identified with them (Elizarenkova 
1972; 1989; Kuz'mina 1976). 

All this leads us to interpret the petroglyphs as ancient sanctuaries. 
Interpreting the semantics of some images and compositions of Andronovo art 
on the basis of Indo-Iranian mythology is a serious argument in favor of setting 
the Aryan homeland in the steppe. 



Some mythological representations 

There exists a vast amount of literature on the ideological beliefs and ritual 
practices of the Indo-Iranians. It is significant for our theme to emphasize only 
those mythological ideas, ritual practices, social structures of the earliest Indians 
and Iranians that are so closely related, including items of vocabulary, that it 
completely excludes the possibility of an early break of Indo-Iranian unity and a 
long separate development of Iranians and Indians. This refutes the hypothesis of 
R. Ghirshman and confronts T. V. Gamkrelidze's and V. V. Ivanov's suggestion 
that the Scythians and Saka left Iran only in the Iron Age, with unbridgeable 
difficulties in explaining their connections with the Finno-Ugrians. In the second 
place, H. Oldenberg showed that in spite of the genetic closeness of religious 
beliefs, the Vedas and Avesta differ considerably, and that in the Avesta many of 
the heroes play opposite roles to their counterparts in the Vedas. In the third 
place, along with the Mazdayasna people of the Avesta and the Veda worship- 
pers there were other peoples in the Indo-Iranian family who had a different 
religious practice; these were the Vratya, in particular, and other Indo-Aryan but 
non-Vedic tribes that are mentioned in the Mahabharata. This is confirmed by 
the non-orthodox mythology of the Nuristani, which has been studied by G. 
Fussman (1977). It follows that we can only expect a unity in the main system of 
ideological beliefs of the separate Indo-Iranian tribes, whereas their religious 
practice differed in details among various tribes of the Indo-Iranian continuum 
which had already split from the homeland. 

The main event of Indo-Iranian religious practice was the sacrifice (of horse, 
bull and sheep) and the cult of the ancestors, that was reflected in sacrifices and 
in the construction of a tumulus. The words 'tumulus' and 'grave' were common 
Indo-Iranian (Bailey 1957). In the burial rite of the Indo-Iranians food offerings 
and the slaughter of animals were extremely significant. Status burials were fol- 
lowed by a horse-sacrifice (Rigveda 10.56). Judging from Vedic tradition, Indo- 
Iranians practiced both cremation and inhumation with the further construction 
of a mound above the grave and sometimes with further fencing (Mandel' shtam 
1968; Elizarenkova 1972; Litvinsky 1972; Gening 1977; Kuz'mina 1985b; 
1986b, and notes). The Iranians were probably bi-ritual in antiquity. Cremation 
is clear from the criticism of this ritual in Yasna 65.8 and in the Videvdat (3.41; 
6). Fire played a great role in the burial ceremony which is reflected in the 
Rigveda hymns directed to the fire-god Agni. The fire was the main conductor of 
sacrifice and it was an intermediary between the human world and the gods. This 
idea is reflected also in the developed hearth and ash cult both in Vedic and in 
Avestan tradition (Hertel 1925). 



1 84 CHAPTER ELEVEN 

The farming cultures of the Indo-Near Eastern region of the 2nd millennium 
BC do not reflect this picture. There in their temples they made libations and 
sacrificed objects, rarely ovicaprids; horse -offerings are unknown. Burials were 
places under the house floors, sometimes in earthen graves; there is no evidence 
for tumuli, stone circles, stelae and wooden houses. 

On the contrary under the tumulus burial rite there was a frame -work in the 
grave; a horse, ram, or bull was slaughtered on the grave or at the funeral feast; 
the warriors were buried together with horses and chariots, and the sacrifices of 
animals at different ceremonies: all these are typical of the culture of the 
Eurasian steppes. 

These characteristic features are found in a number of steppe cultures in the 
2nd millenium BC. A. M. Mandel'shtam demonstrated some specific convergen- 
ces between the Bishkent culture and the Indo-Aryan one and he showed their 
connection with Andronovo rituals. Among a number of specific Indo-Aryan- 
Andronovo correspondences are: the construction of a round or rectangular 
enclosure, a frame-work inside the grave of stone or wood, the dead laying head 
to the west with face to the south towards the kingdom of the dead, and, what is 
more important, cremation (Gonda 1962; Tiwari 1979). The convergence of 
Indo-Aryan ritual with Andronovo and especially with Fedorovo bears a sys- 
temic character and speaks not for just individual traits but a whole constellation 
of characteristics. Thus, the analysis of the burial rite of representatives of some 
of the archaeological cultures of the Eurasian steppes confirms the hypothesis of 
an Indo-Iranian homeland in this zone. The specific complex of correspondences 
between the Bishkent and Andronovo, especially Fedorovo, ritual with Indo- 
Aryan hints at the possible connection of these groups of the Indo-Iranian 
continuum with the Indo-Aryans. 

This hypothesis is supported by the evidence of social stratification of 
Andronovo society in light of the data concerning Indo-Iranian tradition 
(Smirnov and Kuz'mina 1977; Kuz'mina 1985b; 1986b). Rigvedic and Avestan 
terms indicate that there were three social groups recognized in Indo-Iranian 
society (Benveniste 1932; 1933; D'yakonov, M. 1954; 1961; Ivanov and 
Toporov 1960; ITN; Bongard-Levin and Il'in 1969; Grantovsky 1970): priests, 
warriors, and ordinary people. The commoner was called vaisya in the Rigveda 
and vastryo ftuyant in the Avesta. Farmers in Sasanid Iran were called by a 
derivative of this word but its literal translation is 'delivering grass to the cattle'. 
The Avestan vastrya is related to Hittite westara 'shepherd' (Ivanov 1957). The 
warrior in the Avesta was denoted by the term rathaestar i.e. 'standing on a 
chariot'. This term was known also to the Indians although more often they used 
ksatriya 'a warrior possessing power' and rajanya 'regal looking'. The ksatriya 
insignia is a bow according to the Satapatha-Brahmana (5.3.5); according to 
Aitareya-Brahmana it is a chariot, armor, bow and arrows (7.19). Such sets are 
represented in early Timber-grave and Andronovo cemeteries which makes it 
possible to interpret this fact in the light of Indo-Iranian tradition which provides 
another independent method of verification. 



CHAPTER TWELVE 

MORTUARY PRACTICE 

Alongside ceramics mortuary practice is the most important identifying feature 
of an ethnic group. It comprises a time honored custom sanctioned by 
mythological tradition; it is not connected with economic and cultural type; and 
because it can be preserved for a long period of time it does not change in space 
even when an ethnic group has engaged in distant migrations. It is this stability 
of ritual that makes it such an important ethnic indicator. That is why, if our 
attribution of the Andronovans as the bearers of the Indo-Iranian language is 
correct, then all the details of their burial rite which is reconstructed by 
archaeological data should find a correspondence in the rituals reflected in the 
Vedas, Avesta, and in the later tradition of the Mahabharata and Ramayana. 

The burial rite of Indo-Iranians has attracted the attention of scholars of 
ancient literature since the 19th century. Works by W. Caland (1896), S. Gupta 
(1972), Pandey (1976), and J. Tiwari (1979) should be specially mentioned. 
Comparisons of Indo-Iranian tradition with the archaeological evidence of the 
Asian steppes and Central Asia of the 2nd mill. BC has been made by A. M. 
Mandel'shtam (1968) B. A. Litvinsky (1963; 1972; 1981) and E. E. Kuz'mina 
(1986). Special attention has been paid to the study of the horse cult, because it 
constitutes a peculiarity of the ancient Indo-European, especially Indo-Iranian, 
peoples (Koppers 1938; V. V. Ivanov 1974; Kuz'mina 1977; Mallory 1981; 
Maringer 1981; Polome 1994; Renfrew 2000). The works dedicated to the burial 
rites of other Indo-European peoples have been applied as comparative material. 
Extremely important are the data from the syncronous Hittites (Gurney 1952; 
Ardzinba 1982) and the Greeks (Vermeule 1979), as well as generalizing 
researches on Indo-European burial rite (Jones-Bley 1997; 2002; Hansen 1987), 
mythology (Puhvel 1984; 1987; ed. 1970; Dumezil 1966; Mallory 1989, etc.). 

In Indo-Iranian cosmogony sacrifice was the basis of the world order or rta. 
The universe was born from the sacrifice of the first man (Purusa), and the 
primordial horse of the Indians and bull of the Iranians. The elements, plants, 
animals, and man originated from parts of their bodies. A world created from 
sacrifice was maintained by sacrifice: gods in heaven feed on offerings and 
preserve the world order, the ancestral fathers feed on offerings and dwell in 
eternal meadows and pastures of the dead in the upper sky in the realm of Yama, 
the forefather of people who led the way there ("Where our ancient fathers 
passed beyond, there everyone who is born follows, each on his own path", 
Rigveda 10.14.2). The ancestors help their living kinsmen (the well-being of the 
gods and ancestors depends on their generosity, but if the descendents neglect 
their duties they punish them severely). As was the case with the Indians, the 
Iranians also worshiped their ancestors, bringing rich sacrifices to them at the 
funeral and on feast days. 



1 8 6 CHAPTER TWELVE 

In the Mahabharata, in the Bhagavadgita the divine-creator Brahma "having 
created people together with the victim, ordered: Reproduce!" 

— "And sacrificing be sated with your sacrifice 
Gratifying yourself and gratifying the gods 
There will be great help from the sacrifice to you. 
Accepting these sacrifices in the celestial hall 
The pleased gods will reward you." 

In the Rigveda the sacrifice is glorified many times: 

"Offer to Yama the oblation rich in butter, and go forth. So may he intercede 
for us among the gods, so that we may live out a long life-span. 
Offer to Yama, to the king, the oblation rich in honey. We bow down before 
the sages born in the ancient times, the ancient path-makers." (10. 14. 14-15) 

A special hymn is dedicated to the funeral in Atharvaveda 18.4. 

There is a detailed prescription in the Atharvaveda how to perform a sacrifice. It 
is said in hymns sung at the slaughter of a goat (9.5) and cow (9.9) that they must 
be cut with a knife by limbs, "with hymns put the cauldron on the fire" and the 
separated parts of the body are enumerated: head, front and hind legs, skin, ribs, 
each of them substitutes for the whole animal and it will provide the donor 'milk 
curds, milk, melted butter, honey'. And just the same parts of body were brought 
by the ancient Iranians according to the request of Yima to the gods and the 
ancestral Fravashis for a reward. The Fravashis presented a "herd of cattle and 
hordes of sons, a fast horse and solid chariot" to those who "will bow to them 
with meat in hand, clothes in hand, with prayer, worthy of Asa's reward". The 
late Avestan Bundahisn enjoins one to remember the dead by good words 
because they can avert troubles from descendants; this is clarified in the Rigveda 
(10.56.6) where "The fathers have established their own offspring as paternal 
power, like a thread stretched out among those who are to follow." 

The fire-god Agni conveys the sacrifice to the heavens and the deceased 
himself is viewed as a sacrifice in the cremation ritual, where the gods will 
revive the corpse to new life in heaven. The funeral feast also guarantees rebirth. 

In the Mahabharata (1 1.26-29) there is a picturesque description of the burial 
of the Pandavas and Kauravas after their battle. After the fight the surviving 
warriors gathered their own and the alien dead: 

"Highly honored, they gathered dry wood, 
Different weapons, broken chariots, 
On fires carefully led, the main heroes were burnt 
According to posthumous rites" 

and honored them with "sacrifices burnt on blazing fires, with an offering of 
melted butter", performed a funeral feast to the ancestors, and singers glorified 
the deeds of great heroes, and for the entire night the purifying splendid fires 
shone over the battlefield while hymns sounded and women wept (11.26.37-40). 



MORTUARY PRACTICES 187 

The cremated heroes departed to the Otherworld, similar to the worlds of Indra 
(11.26.11) and ascended to the heavens, because honest death in battle "opens 
wide the splendid gates of heaven." 

The burial rites of the Indo-Aryans are found in the other great epic poem, 
the Ramayana. In the sixth book 'Battle', King Ravana's body was covered with 
cloth, put on a couch and his fame was sung, it was carried. In front of the 
procession went those expert in the Yajurveda — the brahmans holding the 
burning torches, who erected an altar, built a fire and placed Ravana's body 
there, sprinkling it with ghee. After a libation to the ancestors and the slaughter 
of sacrificial animals they lit a fire, while women mourned and wept. 

Judging from the Vedic texts and modern ethnographic data in India the 
deceased asked Earth to accept him, looking south to the country of the dead. 
Prayers were read above him, "we go to the gods, to the supreme light" (Pandey 
1982: 191), and his relatives repeated: "Go to the Sun!" The dead was placed to 
the west of the domestic hearth and then the body was carried from the 
settlement on a mortuary vehicle harnessed to two bulls, followed by mourning 
relatives, who walked with their hair let down and did not look back at the 
dwellings of the people, or toward the east, the home of the gods (Pandey 1976; 
1982: 190-194). 

The place of burial was set west or south-west from the settlement and it had 
to be washed by water from the west, from the direction of the setting sun. On 
the burial ground, on a bunch of reeds or mat — a symbol of immortality and 
moisture — they put cult objects and offered sacrifice to Yama, king of the dead. 
The deceased was cleansed, robed in new linen clothes; small pieces of gold, 
another symbol of immortality, were placed on the eyes, nose, and mouth. 

After that, according to the Satapatha Brahmana (12.8.1) they addressed the 
god Pusan, the protector of domestic cattle and roads, with these words: "Pusan 
carries you to the ancestors"; they then slaughtered the sacrificial animal. A 
black cow was given to the brahman (priest), a goat or ram to the vaisya 
(commoner), and his horse and chariot to the ksatriya (warrior); these were to 
help the deceased to cross the river to the kingdom of the dead. It is said in the 
Rigveda, in the burial hymn to a horse (10.56) that carries a man's body to 
heaven: "By merging with a body, grow lovely, dear to the gods in the highest 
birthplace. Victorious racehorse, let your body, carrying a body, bring blessings 
to us and safety to you." 

The horse is burnt with its master. This ritual was also known among the 
Greeks: in the Iliad (23.175) Achilles brings horses to the funeral pyre of his 
friend Patroclus (Vermeule 1979). And in & Rigveda hymn (10.135.3) the dead 
father of the boy goes to Yama's kingdom on "a new chariot without wheels, 
which had only one shaft but can travel in all directions." 

After that followed the gifts and libations to Yama and the ancestors: 

"The one who has passed beyond along the great, steep straits, spying out the 
path for many, the son of Vivas van, the gatherer of men, King Yama — honor 
him with the oblation. 

Yama was the first to find the way for us, this pasture that shall not be taken 
away. Where our ancient fathers passed beyond, there everyone who is born 
follows, each on his own path" {Rigveda 10.14.1). 



1 8 8 CHAPTER TWELVE 

They asked the sacrificed cow that would "give us back the cows, oh Indra" 
Rigveda 10.19.6). 

The burial started after such rites. The early Indians of the Rigveda period 
buried hermits and small children in the earth and even later, inhumation is 
mentioned in the Laws of Manu and is used till this day. But cremation became 
the prevailing ritual. On a stack of firewood covered by reeds they laid the 
deceased. His personal sacrificial utensils, vessels, gold decorations and bow 
"for him to defend us" were put near the ksatriya, his widow was told to: 
"perform the ancient custom", and she lies on the right side of the deceased. 
Then the son is given gold, and instructed to take the "bow from the hand of the 
dead man, to be our supremacy and glory and power" {Rigveda 10.18.9); the 
widow is raised by the younger brother of the dead with the words: "Rise up, 
woman, into the world of the living", and they go round the place prepared for 
the cremation three times from right to left with water: "Let water flow for you 
in three worlds: in this world, in the air and in the sky" (Macdonell 1900: 126). 

After that an animal is sacrificed to the fire -god Agni who bears the deceased 
to the heavens with his flame: "The goat is your share; burn him with your heat! 
. . . carry this man to the world of those who have done good deeds" {Rigveda 
10.16.4). The corpse is covered by the limbs of the slain animal: 

"Gird yourself with the limbs of the cow as an armor against Agni, and cover 
yourself with fat and suet, so that he will not embrace you with his impetuous 
heat in his passionate desire to burn you up" {Rigveda 10. 16.7). 

Three sacrificial fires are lit from the fire taken from the domestic hearth, the 
garhapatya-. The funerary fire is set on three sides while one exclaims: 

"For good, O Agni, burn him from behind, for good from the front! ... 

Let fires enjoy him, sacrifice the whole! 

The one who sacrificed ascended the formed fire, 

Ready to fly in sky" {Atharvaveda 18.4.11-14) 

They address Yama with a prayer. The funeral fire burns for about three hours. 
Then water is poured on it and a new purifying fire is lit, and offerings to the 
gods and ancestors are brought (Caland 1896: 19, 58, 92). Finally the mourning 
concludes. The participants wash themselves and address the Sun: "We have 
returned from darkness", "These living have separated themselves from the 
dead... we have gone towards the dance and laughter" {Atharvaveda 12.2.22). 

The ashes were spread or better thrown into a river, especially the sacred 
Ganges or Jamuna in modern India (Pandey 1982). In ancient time the ashes 
were gathered on a flat dish which could not hold water: "Collect your body in 
joints! By prayer I make parts of your body" {Atharvaveda 17.4.52). Ashes were 
put into the grave, left in the dish or spread on the bottom. Pots with sacrificial 
melted butter were also put in the grave: in the Rigveda, in the hymn to the 
funeral fire (10.16.8) Agni is asked: "do not overturn this cup (the ritual pot 
kumbha - E.K.) that is dear to the gods" and in the funeral hymn in the 
Atharvaveda (18.4.25) pots covered with flat cakes or stones are mentioned "full 
of honey and exuding fat". 



MORTUARY PRACTICES 189 

The rite of inhumation is also described in the Rigveda funeral hymn (10.18, 
verses 10, 11, 14): 

"Creep away to this broad, vast earth... 

Let her guard you from the lap of Destruction. 

Open up, earth; do not crush him. Be easy for him to enter and to burrow in. 

Earth, wrap him up as a mother wraps a son in the edge of her skirt... 

On a day that will come, they will lay me in the earth". 

At both the inhumation of the corpse as well as the cremation the wife lay near 
her husband and then she was told: 

"Rise up, woman, into the world of the living. Come here; you are lying 
beside a man whose life's breath has gone" {Rigveda 10.18.8). 

The deceased was accompanied by parts of sacrificed animals, his personal 
belongings, weapon, vessels; these were probably placed at the head as was the 
receptacle for cremation. 

The grave constructed for the two rites was considered by the Indo-Aryans as 
a house for the dead. 

"Let the earth as she opens up stay firm, for a thousand pillars must be set 
up. . .and let them be a refuge for him for all his days. . . 

Let the fathers hold up this pillar for you; let Yama build a house for you 
here" {Rigveda 10.18.12-13). 

The Atharvaveda funeral hymn is extremely interesting (18.4.37): 

"This frame-work is made by constructing logs 
Kinsmen, come and look at him! 
This mortal goes into eternity. 
Make him a house". 

In his translation B. L. Ogibenin describes the pillar as a 'column, supporting a 
vault'; in the translation of T. Y. Elizarenkova the frame-work is called a 'pile'; 
a wall or fence appears more likely. 

What do we gain from these texts? Kinsmen made a grave as a model of a 
house for the living. It was dug into earth, its walls were strengthened with 
wood, supporting pillars were erected and a frame-work was made from logs. 
Moreover, the same terms were used for house building as those for the 
supporting pillar, the axis mundi. 

After some time (usually ten days) the grave was covered by an earthen 
mound both for a cremation and inhumation: "I shore up the earth all around 
you" {Rigveda 10.14.13), or in the translation by B. L. Ogibenin of Atharvaveda 
18.4: "I deepen you-earth into earth..., cover him, oh earth" and "I set up this 
fence to guard against the world of the dead": 

"I set up this wall for the living. . . 
Let them... bury death in this hill" {Rigveda 10.18.4; Atharvaveda 12.2.22) 



190 CHAPTER TWELVE 

H. Bailey (1957) established that the word 'grave' is common Indo-European 
and in other languages it means 'kurgan', 'tumulus'; the word 'ashes' is also 
common Indo-European. Surrounded by a stone wall the tumulus had either a 
round or rectangular form, but the Satapatha-Brahmana prefers the latter; in 
other cases a round construction is the best. The sraddha-, the funeral feast, was 
held on the third and thirteenth day after the funeral. Flat cakes, milk, fat, honey, 
and water were brought to the deceased and the ancestors: 

"These pots covered by flat cakes, . . . 

Let them be full... of honey, fat exuding 

These grains that I scatter for you, . . . 

Let king Yama approve them for you" {Atharvaveda 18.4.15-26). 

Ancestor sacrifices accompanied all the most important family, calendrical and 
clan celebrations. 

The fence that separated the dead from the living was made of wood or stone. 
It was known not only in the Indo-Aryan but also in other Indo-European 
traditions (Lincoln 1991; Delia Volpe 1992; Jones-Bley 1997; 2002). 

Unfortunately we have no such reliable and early evidence about the Iranian 
practice. Strict regulation of burial rites is preserved only in later parts of the 
Avesta, which were strongly revised by priest -magicians, and which obviously 
did not correspond to the ancient practice. It is possible only to state on the basis 
of some passages and evidence that Iranians, just as the Indians, initially buried 
their dead in the earth (Videvdat 3.41; 5.7, 47, 48; 6.8; Yast 13) or cremated 
them; orthodox Zarathustrians regarded the customs of the Arachosians to bury 
or burn bodies as 'foul' (Videvdat 1) and ordered that corpses be placed in the 
open air to be devoured by birds of prey (Videvdat 5.7, 45, 46). Herodotus 
described this practice in Achaemenid Iran, but he stressed that "however, the 
Persians in general cover the body with wax and then bury it." 

The peoples of Central Asia and neighboring regions: Bactrians, Sogdians, 
Parthians, Caspians, Derbies, Hircanians (Strabo 11.2.9-13) knew the custom of 
displaying the corpse. But this custom was never spread among the Saka nor 
Scythians, nor their descendants, the modern Ossetes, nor the Iranian-speaking 
peoples of mountainous Tadzhikistan and the Hindukush and the nomads of Iran. 
These all bury their dead under a kurgan, often preserving ancient Iranian 
customs, sometimes corresponding to Vedic ritual in every detail. All Iranians 
worshiped Yima (Yama) in antiquity, but he did not become the god of the dead; 
he was a light sunny first man and king, in whose country there was general 
prosperity. The Yima cult was unusually popular in Central Asia and it was 
preserved till the Middle Ages and tales about him can be heard even nowadays. 
Yima taught people to perform animal sacrifices similar to those of the Vedas. 

Two terrible dogs, descendants of the primordial dog Sarama, guard the 
entrance to the kingdom of the dead where Yama rules. The way there is difficult 
and dangerous. It passes over a narrow bridge which the Iranians call the 
Cinvant- 'selected way': good deeds will help the righteous to cross, bad deeds 
will push the unjust into the abyss of an underground ocean where, as the Indians 
thought, lived 'underwater fire', a monster with a horse skull and open mouth 
instead of a head and the serpent of the depths, Ahir budhnyds. 



MORTUARY PRACTICES 1 9 1 

Ideas having much in common with Indo-Iranian exist among other Indo- 
European peoples: there is the celestial Valhalla and underground Hell of the 
Scandinavians, the heavenly Elysian fields and subterranean Tartarus of the 
Greeks, the three-headed dog Cerberus at the entrance of Hades' realm, the 
crossing over the river Styx in Hades' kingdom, and the lord of the sea, 
Poseidon, in the form of a black horse (Puhvel ed. 1970; 1987; Vermeule 1979; 
Jones-Bley 1997; 2002). 

We know about these beliefs from mythological texts: The Eddas, Iliad, the 
works of Hesiod, tragedies, and monuments of art: Scandinavian grave stelae and 
Greek vessels on which solar signs are depicted in the upper sphere: a cross in a 
circle or swastika and birds in the middle, a funeral chariot or ship of the dead, 
and at the bottom — earth symbols or a monster. Another common idea was the 
returning home of hosts of ancestors from time to time, and heroes whose 
victories live eternally in their descendents' memory, inspiring them to new 
deeds. 

Cremation and horse sacrifice are typical only of the Indo-European peoples 
such as the Hittites, judging from the excavation of the Osmankayasy burial of 
the 14th 13th centuries BC. In Iran, however, horse sacrifice appears even later, 
only at the end of the 2nd, beginning of the 1st millennium BC, and it is 
obviously connected with the arrival of the first Iranians, who communicated its 
present name to this country ('Iran' from ancient Indo-Iranian 'Arya'). 

Tumulus burial and horse sacrifice were unknown to the farmers of the 
southern regions of Central Asia, who buried their dead in the 3rd-2nd millennia 
BC in settlements in family burial-vaults, tholoi, under the floors of houses or in 
out-of-the way quarters (Masson and Sarianidi 1972; Askarov 1977). Later there 
appeared burials in the earth in shallow pits or trenches, which were set outside 
the area of the tell (Ganyalin 1956a). 

Where must we look for the sources of Vedic funerary ritual? Their home- 
land was the wide Eurasian steppe. The bearers of the Pit-grave (Yamnaya) 
culture, that developed at the end of the 4th-3rd millennia BC in the steppes 
from the Danube to the Urals, began to bury their dead in graves covered by 
large kurgan mounds. The cult of the horse and bull appeared already among the 
ancestors of the Pit-grave people in the Mariupol' culture: they engraved images 
of honored animals on bone plates, displayed them on the stone scepters of their 
chiefs, and the animals themselves were sacrificed during feasts and funerals by 
dismembering their bodies, separating the head, legs and skin as the Vedic 
Aryans did and in the way Iranian nomads and the peoples of the Pamirs and 
Hindukush still do. 

The origin of the fire cult is also set to the Pit-grave period. The vehicle cult 
appeared at the same time with a funeral vehicle drawn by bulls placed in the 
grave of a particularly honored man, sometimes buried together with a scepter, 
and accompanied by sacrificed bulls and horses (Piggott 1969; Kuz'mina 1971a; 
1974a). In order not to forget the deeds of ancestral heroes, they were glorified in 
hymns, and their images were placed on stone stelae erected on top of the 
kurgan. The ancestral hero is displayed on them girded with a belt and with a 
sharp-ended staff in his hands. It should be remembered here that girding with a 
belt was the most important rite of the Indo-Iranians and befit only the twice- 
born after initiation. With regard to the pointed staff which was used to prod 



192 CHAPTER TWELVE 

cattle and drive bulls in a vehicle, this was presented to Yima by the supreme 
Iranian god Ahura-Mazdah. From that time onwards it was possessed by 
shepherd-kings; the word 'shepherd', 'pastor' is of common Indo-European 
origin and later it began to denote master, king and even god (Ivanov 195 1). 

Thus, already in the Pit -grave culture we come across many prototypes of the 
future culture of the Indo-Iranians. This is the conclusion of E. A. Grantovsky 
(1970); V. Georgiev (1956), M. M. D'yakonov (1961); G M. Bongard-Levin et 
al. (1969) who saw the distant ancestors of the Indo-Iranians in the tribes of the 
Pit-grave culture. N. Ya. Merpert (1960) came to the same conclusion indepen- 
dently, applying the retrospective archaeological method. 

It has been said that the kurgan rite and animal sacrifice, including the horse 
which was much honored by the Indo-Aryans, existed among different peoples 
of the Eurasian steppe. But only in one culture, in the Andronovo, do we find a 
combination of all the specific elements of the Vedic funeral complex. How did 
the Andronovans bury their dead? 

Like the Indo-Aryans they buried their dead far from their settlement, on the 
river bank, and they oriented the graves to the west or south-west, the direction 
of the setting sun. They dug the graves into the earth and constructed a timber 
frame -work; in the treeless regions they replaced it with a stone cist. They 
erected overhead a cover on the central supporting pillar. The grave was covered 
with earth by a kurgan of pyramidal form; it was surrounded by a circular or 
square stone fence. Like the Vedic Aryans and earliest Iranians they too knew 
both burial rites — cremation and inhumation. The Andronovo-Fedorovo type 
practised cremation. It is important to take into account the directions of Sanskrit 
commentators that the Indo-Aryans gathered ashes into a flat vessel, and among 
all the Eurasian cultures such ash-dishes were known only in the Andronovo 
culture. In the first case we find a heap of scattered ashes in a grave and very 
often there is a clay dish in which it was carried. There are also horse ribs and 
some vessels at the wall. 

Neither weapon, nor ornaments, nor bones of other sacrificial animals are 
found. But in light of the Indo-Aryan data this was according to ritual 
prescription. Objects were taken away, and the sacrificial animal was burnt on a 
funeral fire. 

Inhumation dominated among the other Andronovo (Sintashta, Petrovka and 
following Alakul') types. 

The essence and many details of Andronovo rituals can be explained only in 
the light of Indo-Iranian literary data. The Indo-Iranians conceived of the earth as 
a square or circle, or these figures were inscribed into one another. A round 
kurgan with a square grave in the center was a microcosm of the universe. 

Each part of the world had its own god-protector. In India Indra was protector 
of the east, Varuna the west, Soma the north, and Yama protected the south. The 
orientation according to parts of the world was extremely important, but varied 
among different peoples: for the Indians the bad, dangerous side was south; for 
the Iranians it was the north. This complex of ideas probably preconditioned the 
strict orientation of graves in a single direction which became a traditional 
characteristic of the culture of an actual ethnic group. 

In Sintashta-Petrovka and Alakul' graves the deceased were flexed on the 
side. For a long time this was explained by the idea of death as a dream and so 



MORTUARY PRACTICES 193 

the dead were placed in the pose of sleep or because ancient people were afraid 
of the dead and tied them tightly by ropes to prevent them from walking. But 
Vedic texts explicitly state that the earth provides a rest for the deceased as a 
mother does for her son and that he will survive to a new life; thus the flexed 
position mirrors the foetal position. To achieve the necessary flexed pose the 
deceased was really tied, and the god of death Yama carried a rope as his own 
attribute, and he was constantly called 'tying the dead'. 

There are traces of fire rituals in Petrovka burials: ashes strewn on the floor 
of the grave, coals and ashes in the mound, and burning of the ground above the 
grave cover. A quite similar picture is witnessed in Early Timber-grave burials. 
Although it is all a manifestation of a single fire cult the presence of fire is the 
result of quite different ritual actions, conducted at different times and for differ- 
ent reasons: the ashes in a Fedorovo grave are associated with the cremation of 
the deceased, the ashes in Petrovka and Timber-grave burials where we find 
inhumations are probably the residue of an animal sacrificed to Agni or Yama, a 
burnt wooden ceiling or burnt clay ground above it may be compared to the 
purifying fires set by the Indo-Aryans after funerals. 

There are usually two to three vessels in Andronovo graves. They are at the 
head of the dead, just as it is described in the Sanskrit texts. Parts of sacrificial 
animals were also placed in Alakul' graves: the skull and legs of a horse, bull or 
ram. Sometimes vessels and bones were placed above the grave pit and at the 
edge of the fence. For a long time scholars believed that this was a food offering 
given to the dead for the afterlife. Examination of the original texts reveals that 
the ancient Indo-Aryans thought quite differently. In reality, remains found 
under the kurgan mound are traces of not one but of several rituals conducted at 
different times and for different purposes. In the first place, sacrifice was given 
not to the dead or for the dead but for all the deceased's ancestors headed by 
Yama. They are called upon many times to descend to the earth, sit on the strewn 
straw and taste the food prepared for them. They are reminded not to forget to 
thank the sacrificers and to send them cattle and healthy posterity as a reward. In 
the second place, some sacrifices were brought in a certain order to the 
individual gods — Yama, Agni, and Pusan, and the texts indicate for whom and 
for what reason the animal has been presented. In the third place, the main 
sacrificial animal was not there to please the dead, ancestors or gods but as a 
means of transport. It is said directly in the texts that the ram or bull is the animal 
of the protector of roads, Pusan, and they will help the deceased to cross to the 
kingdom of the dead. A special Rigvedic hymn was sung in honor of the horse 
that will carry its master to the sky. It is certain that such chariot horses accom- 
panied Sintashta-Petrovka and Alakul' burials. Moreover, they lay in pairs as if 
in harness and they were sometimes put not into the burial but on its roof. In 
these cemeteries there are chariot burials. In some cases the chariot was probably 
put in the grave with the wheels removed and without the shaft as in a Rigvedic 
hymn where the boy talks to his dead father. In other cases the imprints of 
wheels indicate that the chariots were put into special grooves to a depth of 
approximately one third of the wheel diameter. This detail is not present in the 
funeral hymn, but in texts dealing with construction, it is prescribed to build 
special sheds for chariots near a house and it is recommended to prepare grooves 



194 CHAPTER TWELVE 

for the wheels of the chariot. One can observe a striking resemblance in the 
smallest details. The same grooves are fixed in Chinese che-ma-kens for chariots. 

Finally, in Andronovo kurgans vessels and animal bones were placed not 
only in graves, but also on the edge of the mound, near the stones of the 
enclosure. In the light of Vedic data this complex was a funeral feast conducted 
some time after the funeral. There are only skulls and legs of animals separated 
into joints in strict prescription with sacrificial immolation in the Atharvaveda. It 
is known that according to the beliefs of all Indo-Europeans a collective feast or 
funeral feast guarantees future revival. The funeral feast and collective meal are 
the most important rituals both in the calendar and marriage and funeral 
ceremonies. We can see this in the fact that Russian has a word of elevated 
register, zhrets (priest) and a word of low style zhrat' (to gobble) which are 
derived from the same root; funeral feasts are also depicted on paintings in Greek 
tombs and sarcophagi. 

Vedic data can explain another peculiarity of the Andronovo burial rite: 
paired burials of different sexes. Approximately 2% of Andronovo graves are 
paired burials of man and woman. They lie facing each other, man on the left, 
woman on the right side. 

These burials have been intriguing scientists for a long time. V. S. Sorokin 
(1962) and M. A. Itina (1977) believed that the woman was a forcibly buried 
slave or concubine of the man. This conclusion received universal recognition, 
although there is a serious reason to doubt it. 

In the first book of the Mahabharata (91-100) it is said that when the 
forefather of the Pandavas, King Pandu, died, his young wife Madrl, mother of 
the twins Nakula and Sahadeva, voluntarily mounted the funeral pyre after her 
husband. And in the third book (277-283) when Satyavan was doomed to death 
by the gods, the all-ruling Yama came with a rope to tie him. But his wife 
Savitrl, who never parted with her husband, was ready to go with him after 
Yama: 

"Let wife follow her husband everywhere. 

I was with him when he lived, I will be with him when he died." 

It is said in the Laws of Manu that after her husband's death a woman was free to 
chose whether or not she must follow him to the Otherworld. 

Judging from the Mahabharata, Atharvaveda and Rigveda this custom of sati 
was never strictly observed and suicide was replaced by a symbolic act — the 
widow only lay down on the right side of the dead husband, exactly as in 
Andronovo burials. 

Iranian-speaking peoples transformed this rite differently. There a widow 
placed a cut plait of her hair. This is done to the present time, for example, 
among Kurdish women and this custom was also found not long ago in Ossetia 
(Kaloev 1971). The custom of the voluntary death of a widow to be buried with 
her husband was known to other Indo-European peoples: Herodotus says (4) that 
so did the wives of noble Thracians, neighbors of the Iranian Scythians, and the 
same rite existed among Scandinavians, as is described in the sagas. The reason 
for sati was that the women could enter Valhalla only at the price of a sacrificial 
death. This means that it was not a slave woman but the full loving wife who 



MORTUARY PRACTICES 195 

followed her husband. This conclusion is also very important for understanding 
family relations in Andronovo society as it shows that the extended family living 
under a single roof consisted of small nuclear families (either son with spouse or 
daughter with spouse). The small number of family burials supports this 
supposition. In the Baytu burial, for example, we discovered a grave where the 
husband and wife lay and a child was buried at their feet. 

More numerous in the Andronovo culture were burials of a mother with one, 
sometimes two, small children placed near her in one grave or buried closely in a 
special pit. These are rather difficult to explain. It is generally accepted that the 
majority of early peoples of the world strictly observed exogamy where the wife 
was necessarily taken from another kinship line. Then the husband and children 
would be buried in the cemetery of his kin, while his wife would be brought to 
the cemetery of her kin. But if women are buried together with children then the 
Andronovans did not observe exogamy; they had endogamy where marriages 
were made within kin groups. Such a supposition is affirmed by another category 
of Andronovo material culture: ceramics. It has already been said that in every 
micro-district and even in every settlement ceramics have specific features, and 
it is possible only when a daughter trained by her mother has remained in her 
native settlement. It means that the Andronovans had matrilocal marriages, i.e., 
the husband came to the wife's settlement or he was from the same kin. 

This important demographic conclusion can also be used to affirm the Indo- 
Iranian attribution of the Andronovans because both Indians and Iranians did not 
know the custom of exogamy in antiquity (Herzfeld 1941, 1: 119; Livshits 1963: 
506). Kinsmen marry each other among modern Iranian peoples, e.g., Kurds, 
Luri, mountain Tadzhiks and others (Andreev 1927; 1958). This ancient custom 
could be attributed to the caste system in India when marriage was within a caste 
without taking into account kinship affiliation. 

In the Andronovo culture apart from child burial in a clan cemetery there 
were special cemeteries where infants were buried. And in the Indie Laws of 
Manu (5.68, 69) it is said that a "child who died before two years must be carried 
by relatives from the settlement, covered with flowers, and buried in the pure 
earth without gathering its bones afterwards. Such a child must not be burnt on a 
pyre and water must not be poured over it." 

Thus, the variety of Andronovo funeral rites finds a complete and thorough 
correspondence in early Indie texts. Survivals of these rites are still present 
among the modern peoples of India, Iran, the Hindukush, Pamirs, and Ossetia. 

This analysis shows that the whole complex of Andronovo burial rites finds 
analogies among other Indo-European peoples. The kurgan with the surrounding 
fence and the burial construction beneath it represents the house of the dead. The 
combination of cremation and inhumation, sati, the rite of sacrificing a pair of 
horses and a chariot as a means of transport, sacrifice of a cow, sheep and ram, 
sometimes the burial of a dog, the placement of vessels with ritual food in a 
grave (honey, milk, fat), as well as weapons and ornaments characterize the 
burial practice of the majority of Indo-European peoples: Hittites, Greeks, 
Germans, Baits, Slavs, etc. It leads to the undisputable statement that the 
Andronovans were Indo-Europeans. However, the common Indo-European 
character of the whole burial complex does not, strictly speaking, permit one to 
declare the Andronovans as Indo-Aryans. V. F. Gening (1977) erred when he 



196 CHAPTER TWELVE 

declared that the Sintashta people were Indo-Aryans when he only compared the 
burial rites with those of common Indo-European, not the rites specific to the 
Indo-Aryans themselves. For this reason L. A. Lelekov (1982) opposed him. The 
Indo-Aryan attribution of the Andronovans can only be established on the basis 
of the similarity between Andronovo and specifically Vedic burial ritual. 

In accordance with my methodology the decisive arguments for the ethnic 
attribution of the Andronovo culture rests on the evidence of material culture, 
economy and cultural complex. The data on the burial rite and beliefs can be 
used only to verify the hypothesis. As the Andronovo burial rite is in full 
systematic agreement with the Vedic one, this fact can be used as a decisive 
argument for declaring the Andronovans as Indo-Iranians. 

The absence of a pig sacrifice and the special role of the horse and chariot 
serve as particular ethnic indicators that distinguish Indo-Iranians from other 
Indo-European peoples and single out the Andronovans from neighboring tribes. 
The rite of cremation which is characteristic only of the Fedorovo tribes of the 
Andronovo culture makes this population close to the Indo-Aryans. 

Another important argument in favor of setting the Indo-Iranian homeland in 
the steppe is the widespread idea of an afterlife as a heavenly pasture which is 
reflected in Vedic literature, and which is close to the Greek Elysian Fields. It 
has been thoroughly studied by Puhvel (1969). The mythologeme of an afterlife 
as a wide heavenly pasture could have appeared only among the herdsmen 
people of the steppes. 

If we recall that burial rite is the most important ethnographic feature of a 
living people and that it can be preserved even over long development in one 
place and after migration into another economic and geographical zone, then this 
correspondence is the safest basis for acknowledging the Andronovans as Indo- 
Iranians. 

This conclusion, important in itself, has great importance also because we 
can use Indo-Iranian data to reconstruct the social structure of Andronovo 
society. As mentioned above, they led a pastoral way of life, did not know 
specialized crafts, lived in large houses in large family communities; moreover, 
while analyzing materials from settlements there are seemingly no traces of 
social and economic differentiation. But is this really so? The study of 
Andronovo cemeteries yields different picture. Beginning with the monuments 
of the Sintashta-Petrovka type of the 18 th 16th centuries BC it is possible to see 
the considerable differences in size of constructions and grave-goods from the 
various kurgans. There is a notion in the archaeological literature of a 'standard 
set of a burial' (Alekshin 1981; 1986b), i.e., a complex of the most typical chara- 
cteristics of given culture, that include the size of the kurgan and grave pit, the 
complexity of construction, the number of vessels, sacrificial animals, and grave 
goods. 

In the Sintashta-Petrovka type cemeteries with an average grave size of 2.5- 
3x2m and 1.5-2m depth, there are graves measuring 4x3m, to 3.7m deep. Such 
graves do contain the most complex wooden constructions, the largest number of 
sacrificial animals, the best modeled vessels with the richest ornament. Chariots 
and harnessed horse burials are found in these graves; near the deceased there is 
a set of weapons, including a knife, stone and bronze arrows, stone mace, 
spearhead, and chisel (Tkachev 1995; 1996; Epimakhov 1995; 1996; 2002). 



MORTUARY PRACTICES 



197 



The same picture is seen in Alakul' burials of the 15th 14th centuries BC: 
there is an absolute predominance of kurgans from 7m to 20m in diameter, 0.15- 
0.6m high with graves 2-3x1 .4-1 .75m, 0.5-1 .4m deep with a pair of vessels. But 
in every cemetery there are several enormous kurgans up to 40m in diameter, up 
to 1.5m high or constructions surrounded by double and even triple rings of 
monumental stone slabs. There are graves in such kurgans that measure 
3. 2x2. 5m, 2.3m deep; their walls were covered with a frame-work of large logs, 
covered by a layer of logs which is sometimes double. These graves were 
unfortunately robbed in antiquity but there we still find numerous bones of 
animals, gold ornaments dropped by robbers, bronze knives, maces, horse cheek- 
pieces, sets of arrows, all of which prove the former richness of the burial. (The 
same differences in the sizes of the kurgans and graves are found in early 
Fedorovo burials, but the absence here of corresponding grave goods diminishes 
our results). One conclusion must be evident: Andronovo society was not 
uniform; there was a group of people who occupied a privileged position. 



Site type 


Construction size 




Average 


Maximum 




Diameter 


Height 


Diameter 


Height 


Late Alakul' 


- 


- 


- 


- 


Late 
Kozhumberdy 


2-5 


Enclosure 


8 


- 


Atasu 




Long fences 




- 


Kozhumberdy 


7-10 


0.8 


17 


0.8 


Amangel'dy 


8-12 


0.2-0.7 


16 


0.9 


Early Alakul' 


7-20 


0.15-0.6 


40 


1.5 


Petrovka 


12-20 


0.2-0.4 


- 


- 



Site type 


Pit size 




Average depth 


Maximum depth 


Late Alakul' 


1.35/1.5x0.6/1 


? 


1.6x1.25 


9 


Late 
Kozhumberdy 


1.4/1.5x0.6/0.75 


0.3-0.7 




- 


Atasu 


1.5/1.7x0.8/1 


0.6-1.2 


2.2x1.6 


1.6 


Kozhumberdy 


1.7x0.9/1.1 


0.7/1 


2.9x2 


1.6 


Amangel'dy 


2.2/2.5x1.5/1.75 


0.3/1 


3.2x2 


1.5 


Early Alakul' 


2.3x1.4/1.75 


0.5/1.4 


3.2x2.5 


2.3 


Petrovka 


2.5/5x2/3 


1.5/2 


4x3 


3.7 



Table 12: Evolution of burial constructions and pit dimensions 



CHAPTER THIRTEEN 
INDO-IRANIAN CONTACTS WITH OTHER LINGUISTIC GROUPS 

The study of contacts between peoples belonging to different language families 
is important not only for the reconstruction of ancient cultural relations but also 
for elucidating the homeland of the ethnic groups in contact. The methodology 
requires comparative analysis of three different sources: linguistic, archaeo- 
logical and mythological. 

Indo-Iranians and Finno- Ugrians 

The hypothesis of the relationship between the Hungarians and the Ob-Ugrians 
and their Siberian homeland was suggested already in the 15th century by the 
Italian humanist Enea Silvia Piccolomini (later Pope Pius II). In the 17th century 
the great German scholar Gotfried Leibnitz singled out a Finno-Ugrian language 
unity and marked its borders. The Finno-Ugrian languages comprise four bran- 
ches: Balto-Finnic which spread very early to the west; Volgaic (Mari, Mordvin- 
ic and the now extinct Merya, Muromian, Meshcherian); Permic (Udmurt, 
Komi); Ugrian (Khanty, Mansi, Hungarian); Saami; the Samoyedic branch of the 
Uralic language family are not discussed here. N. Anderson first noticed the 
Indo-European connections with the Finno-Ugrian languages, and B. Munkacsi 
and B. Collinder (1954) raised the issue of borrowings from the Indo-Iranian 
languages. The division into west (Finno-Permic) and east (Ugrian) is assigned 
to the transition between the 3rd and 2nd millennia BC (Fodor 1976; Hajdu 
1985; Redei 1997, Napol'skikh 1997). Numerous data on Indo-Iranian borrow- 
ings in the Finno-Ugrian languages were systematized by K. Redei (1986, 1997), 
E. Korenchy (1972), A. Yoki (1973) and augmented by A. Csillagy (1974), etc. 
Important contributions were made by the Iranists Th. Burrow (1976), V. I. 
Abaev (1972), J. Harmatta (1981), A. V. Lushnikova (1990), and E. Helimsky 
(1996). They all suggested a chronological stratification of the linguistic material 
and proved that contacts were carried out not only with Iranians but also with 
Indo-Aryans and that they go back to deep antiquity. This was also confirmed by 
A. Parpola (1988; see also Carpelan and Parpola 2001). 

Another important source for the study of Finno-Ugrian, Ugrian and Indo- 
Iranian contacts in particular, are the common mythological features of these 
peoples, investigated by B. Munkacsi, K. Karyalainen and A. Kannisto, also by 
Hungarian scholars {Popular Beliefs 1968; Ancient Cultures 1976), Finnish and 
Russian scholars (Aykhenval'd, Petrukhin and Khelimsky 1982; Gertsenberg 
1975; Steblin-Kamensky 1995). Works by V. N. Toporov (1975; 1981) and G 
M. Bongard-Levin and E. A. Grantovsky (1976) are also of particular interest. 

The earliest layer of Indo-Iranian borrowings is defined by common Indo- 
Iranian, Indo-Aryan and proto-Iranian terms relating to three spheres of culture: 
productive economy, social relations and religious beliefs. Words pertaining to 



200 CHAPTER THIRTEEN 

the productive economy include domestic animal names ('sheep', 'ram', 
'Bactrian camel', 'stallion', 'colt', 'piglet', 'calf); terms connected with stock- 
raising products or processing ('udder', 'skin', 'wool', 'cloth', 'spinner'); 
farming terms ('grain', 'awn', 'beer', 'sickle'); names of tools ('awl', 'whip', 
'horn', 'hammer' or 'mace'), probably the word for 'ladder' (or 'bridge'), and 
finally, 'metal (ore)'. A large group of words reflects personal, family and social 
relations (e.g., 'man', 'sister', 'orphan', 'name'), important Indo-Iranian social 
terms, e.g., dasa ('non-Aryan, alien, slave') and asura ('god, rich, master, hero'). 
The word 'price' and words for numerals illustrate the establishment of exchange 
relations. Finally, a considerable part of the borrowed words reflects religious 
beliefs and practice: 'heaven', 'below (lower world)', 'god (happiness)', vajra 
(Indra's weapon), 'dead (mortal)', 'kidney' (organ of the body used in the Aryan 
burial rite) and names of stimulating drinks used both by Indo-Iranian priests and 
by Finno-Ugrian shamans, e.g., 'honey/mead', 'hemp', 'flyagaric'. 

We may ask when and where could contacts occur between Indo-Iranians and 
Finno-Ugrian peoples where the latter accepted words and the skills of a produc- 
tive economy, incipient social stratification, and a complex mythological system 
and developed ritual? 

Decisive arguments for establishing the Finno-Ugrian homeland are the 
names for common plants and animals, which localize the territory of formation 
before its collapse to the forest zone of Euraia, excluding the far north (Cherne- 
tsov 1953, 1963; Osnovy Finno-Ugorskogo yazykoznaniya 1954; Veresh 1984- 
1985). F. P. Koppen, V. V. Radlov, and P. Hajdu (1985) initially placed it west 
of the Urals because of common words for 'elm', 'lime', 'ox', 'pine', 'bee', and 
'honey'. Palynological research on archaeological sites of the 2nd millennium 
BC in the forest and forest-steppe zones of the Urals and western Siberia have 
recovered elm, lime, birch, aspen, and Siberian pine (G. N. Lisitsyna analyzed 
materials from the author's excavations in the southern Urals). It is known that 
wild bees are found in the Urals and in southern Siberia, where bronze objects, 
cast in wax molds, were found. This evidence convinces us of the view ex- 
pressed by the majority of scholars, that the Finno-Ugrian homeland was situated 
in the Urals and southern Siberia. V. V. Napol'skikh (1997) provides the most 
detailed argument for a Uralic homeland that is demonstrated by palynological 
data. It is confirmed by common terms in the Samoyedic languages and 
borrowings from Paleosiberian languages (Helimsky 1988; Nikolaeva 1988). 

In Siberia the beginnings of the productive economy, acquaintance with 
horse-breeding, and developed bronze metallurgy appear in the aboriginal 
settlements and in the cemeteries of Rostovka, Preobrazhenka, Samus' IV, Kro- 
khalevka, etc., which are dated to the 14th— 13th centuries BC and earlier (Ma- 
tyushchenko 1988; Molodin 1973; 1977; 1985). The discovery of Andronovo 
ceramics of the Fedorovo type, metal objects specific to the Andronovo-Fedo- 
rovo metallurgy of eastern Kazakhstan and Semirech'e, which are decorated with 
images of animals found in the mountain-steppe regions of the Tian-Shan and 
Altai, are sufficient to suggest that Andronovo-Fedorovo tribes influenced their 
northern neighbors in the formation of the productive economy, horse-breeding 
and metallurgy on the southern taiga. A more powerful southern impulse from 
the Fedorovo type is evident in another group of sites of the pre -taiga zones of 
the Urals and Siberia. These comprise the following cultures: Cherkasul' (an- 



INDO-IRANIAN CONTACTS WITH OTHER LINGUISTIC GROUPS 20 1 

other name is Zamaraevo and Mezhovka), Pakhomovo, Suzgun, Chernoozer'e, 
and Elovka, that are united under a single name 'Andronovoid' (Gening and 
Eshchenko 1973; Matyushchenko 1973a; 1973b; Kosarev 1974; 1981; Molodin 
1985; Korochkova 1987; Obydennov 1986; 1997; Obydennov and Shorin 1995; 
Potemkina et al. 1995, Shorin 1988; Gening and Stefanov 1993; Gening and 
Stefanova 1994). Although they maintain their forest image with a predominant 
reliance on fishing and hunting as the main basis of the economy, they also 
indicate an acquaintance with incipient farming and stockraising (horse, cattle, 
and ovicaprid bones were found in settlements). By all the most ethnically sen- 
sitive features (domestic architecture, ceramic types, burial rite) they preserve 
the traditions of the preceding aboriginal cultures of the Eneolithic and Early 
Bronze Age, but stock-raising and bronze metallurgy point to the influence of 
Andronovo-Fedorovo populations. The forest zone ceramics were ornamented: 
the decoration reproduces the Fedorovo type although transformed (that is why 
this culture was called Andronovoid). 

Scholars unanimously acknowledge this Andronovoid complex but differ in 
explaining its origin and ethnic attribution. V. I, Moshinskaya (1957: 134) and 
K. V. Sal'nikov (1964a: 22) suggested that the similarity in ceramics between 
the Andronovoid cultures of the forest zone and Andronovo resulted from the 
fact that both were formed on the basis of related Neolithic and Eneolithic cul- 
tures of the Urals and western Siberia. V. N. Chernetsov (1948: 151-153, table 6; 
1951: 29; 1953: 61) traced the development of Fedorovo ornament from the 
Bronze Age to modern Ob-Ugrian decoration, although modern Ugrians employ 
Andronovoid ornament not on ceramics, but in dress and birch-bark (Chernetsov 
1948: 139, Ivanov 1963: 161, fig. 100). Such a conclusion is still accepted with- 
out criticism (Veresh 1978; Potemkina 1983). M. F. Kosarev (1974: 149-151) 
also believed that the Fedorovo and Andronovoid Suzgun and Cherkaskul' sites 
in the Trans Urals were formed on the single base of "the development of the 
tradition of executing comb geometric decoration with a comb in the forest Trans 
Urals", but the appearance of the Andronovoid Molchanovka and Elovka cul- 
tures in the Ob region he explained by Andronovan migration and its influence 
on the Siberian aboriginals. He concluded that the Fedorovo culture was formed 
in the Urals and assigned it to the Finno-Ugrians. However, he stressed that "the 
assumption of an Andronovan (Fedorovo) and Elovka ethnic unity is far from 
indisputable" (Kosarev 1974: 157). V. N. Chernetsov' s hypothesis must be rejec- 
ted at present. Fedorovo monuments are discovered not only in the Urals but also 
in the south of Central Asia and Afghanistan, where Ugrians have never lived 
(see Chapter 3). 

1) The hypothesis of an origin of the Fedorovo type in the Urals has been 
disputed. The sources for Fedorovo ceramic technology and triangular 
ornamentation are found in the Eneolithic of central and eastern Kazakhstan. 

2) Although S. V. Ivanov's (1964: 1-7) conclusion is based on much 
ethnographic material, the Andronovoid ornament complex is characteristic of 
the art not only of the Ob-Ugrians but also of peoples from other language 
families — Kets, Evenks, Dolgans, Yukaghirs and later Turkic-speaking Yakuts 
and Buryats. The Ugrian ornament complex includes: a) the simplest elements 
that originated convergently among different peoples of the world; b) elements 
characteristic of some Eurasian cultures including the Andronovo and Timber- 



202 CHAPTER THIRTEEN 

grave, and continued in the art of the Eurasian steppe peoples; c) specifically 
Ugrian elements (Ivanov 1963: 154-158). 

It follows that survivals of the Andronovo ornamental tradition among 
Uralic-Samoyedic peoples do not prove that the Fedorovo people spoke an 
Ugrian language; it only points to the active and prolonged influence of 
Andronovans on both Finno-Ugrians and other Siberian peoples. 

This deduction is in agreement with the conclusions of linguists that some 
terms connected with the productive economy in the Finno-Ugrian languages (as 
well as the Ket and Turkic language) are borrowed from Indo-European, first of 
all from Indo-Iranian and later Iranian (Joki 1973). As linguists set the Finno- 
Ugrian homeland in the Eurasian forest zone, where the productive economy 
appears in the second half of the 2nd millennium BC, i.e., the period of the 
spread of Andronovoid ceramics, the latter probably reflects Finno-Ugrian and 
other Siberian peoples' perception of Indo-Iranian cultural influences; this 
proves that the Fedorovo people belonged not to Ugrian but to Indo-European, 
most likely Indo-Iranian peoples, to some group that differs from Alakul' but is 
related to it. It must be mentioned that in the given Andronovoid ornamental 
complex there are specific Fedorovo, Alakul' and Kozhumberdy elements 
(Moshinskaya 1957: 124; Ivanov 1963: 161, fig. 100; Matyushchenko 1974; 
Kosarev 1974: fig. 32). This points to the influence of not only Fedorovo but 
other Andronovo tribes as well. 

It is important for ethnogenetic reconstruction that the Fedorovo ornamental 
complex is present in the decoration of modern Iranian-speaking Tadzhiks 
(Bobrinsky 1900) and it is known in Hindustan where it has no sources in the 
local Eneolithic cultures and could spread only with the arrival of the Indo- 
Aryans. This supposition is supported by the evidence of handmade pottery in 
the more traditional districts of north-west Hindustan and textual materials on 
such technology spread by the Vedic Aryans. 

If we identify the Andronovo tribes as Indo-Iranians then we can see the 
Andronovoid cultures as representatives of the Ugrians, and the boundary 
between the forest and steppe zones was the area of their contact with one 
another. Thus, the correlation of evidence on the ethnic formation of two 
different language families strengthens both independent hypotheses about the 
natural habitats of their locations. 

An important argument in favor of both hypotheses is toponymies: the names 
of the sacred Ural mountains — Ripa — is fixed in both ancient Iranian and Ugrian 
traditions; the idea connected with a sacred Meru mountain is found in India and 
among the Finno-Ugrians; and also the common name of the sacred river 
Volga — Ra (Indian Rasa, Avestan Rayha), the sources of which were the Kama 
and the Belaya rivers, flowing from the Urals (Grantovsky 1960; 1998; Bongard- 
Levin and Grantovsky 1983; Abaev 1972; Dovatur and Kallistova et al. 1982; 
Chlenova 1983a;1989; Marquart 1938; Christensten 1943). 

Mythological data provide a third group of sources supporting this hypo- 
thesis. The formation of the Ugrian pantheon and the reflection in it of the cult of 
the three earliest Indo-Iranian gods, evidenced by Indo-Aryans in the 14th centu- 
ry BC in the Near East and among the Indians in the Rigveda and the Iranians in 
the Avesta. This reveals an acquaintance with Indra as shown by the borrowing 
of his sacred weapon, the vajra; there is Varuna, frequently compared to the sup- 



INDO-IRANIAN CONTACTS WITH OTHER LINGUISTIC GROUPS 203 

reme god Numi-Torum in terms of functions and attributes: horse-drawn chariot; 
and Mithra, whose functions and common Iranian epithet 'world contemplating 
man' was adopted by Mir-Susne Hum, who is related (as well as his epithets) 
with a shining white horse. The connection with the horse, whose sacrificed 
burials and images in plastic art appear in the taiga zone on sites exhibiting 
Andronovo influence, proves that horse-raising was borrowed by the Finno- 
Ugrians along with the mythological ideas and rituals similar to the Indian 
asvamedha (Moshinskaya 1978; Kuz'mina 1977). The formation of the image of 
Numi-Torum on a chariot can be safely dated to the 2nd millennium BC when 
chariots were used by the Andronovans in the steppes, as in the Scythian period 
chariot battle tactics were no longer used. 

Other pantheon heroes such as Sornipos ('golden ray') and King Yima also 
demonstrate Ugro-Aryan contacts (Steblin-Kamensky 1995). Analysis of the im- 
age of the Mother-Goddess is of special interest. In Ugrian tradition she is refer- 
red to as a she-beaver and is depicted with two beavers at her sides or standing 
on a beaver (Skalon 1951; Kuz'mina 1988b; 1990; Chlenova 1989). This conta- 
mination between Indo-Iranian and Finno-Ugrian is connected with 1) the aphro- 
disiac power of beaver's secretion providing fertility; 2) the connection between 
the beaver and the two spheres of the Universe — the earth and underground/ 
underwater — related to the mother-earth. It was assumed that the Mother- 
Goddess image associated with the beaver was borrowed from Ugrian peoples by 
ancient Iranians. In Yast 5 of the Avesta, dedicated to ArsdvT Sura Anahita, the 
'stainless' goddess, she is dressed in a beaver coat. It is evident that such an attri- 
bute could appear only in antiquity when the beaver was first an embodiment, 
then a companion, of the goddess, and only in a zone where beavers lived. Anal- 
ysis of osteological materials from the Andronovo settlements of Kipel', Novo- 
Burino, Alekseevka, Shortandy-Bulak, Ust'-Narym, and Malokrasnoyarka 
showed that in the Bronze Age beavers dwelt over the entire Andronovo area 
(Kuz'mina 1988c: 56-57; Afanas'eva 1960; Kozhamkulova 1969), although at 
present this animal is not found in Kazakhstan. 

The beaver's name in the Indo-Iranian languages is common and refers to a 
water animal with sparkling brown skin — both beaver and otter as in the Ugrian 
languages (Gamkrelidze and Ivanov 1984: 2: 529-531). Correlation of the 
mother-goddess with the sphere of rivers and water animals is characteristic of 
all Iranian-speaking peoples: among Iranians she is protectress of the water, 
ArsdvT Sura Anahita in beaver furcoat; among the Scythians, the ancestral 
mother, daughter of the Dnieper (Don Apris), a snakefooted nymph. In Greek art 
she is depicted with sea-horses, among the Ossetes she is the ancestress 
Dzherassy living in the depths of the water, the daughter of the water ruler Don 
Betyr and she is depicted as a turtle (Kuz'mina 1977; 1988c). 

Probably the ideas about a world axis and the Sampo mill, the spirit of the 
wind, retained only in Finnish tradition, was connected with the earliest Indo- 
Iranian god of the wind, Vata. Cults of Kara, a sacred fish of the Volga river, 
found in Ugrian tradition, and the fantastic elk Sarabha were, according to Th. 
Burrow, G M. Bongard-Levin and E. A. Grantovsky, adopted by the Indo- 
Iranians from the Finno-Ugrians and they can be referred back to the same 
ancient period. The name of the Bactrian camel and its cult is an ancient Finno- 
Ugrian borrowing from Indo-Iranian (Kuz'mina 1963a). The practice of the 



204 CHAPTER THIRTEEN 

shamans also has Indo-Iranian sources. The open clothing of the Khanty 
(Prytkova 1953) that does not correspond to a northern ecology is an Andronovo 
borrowing. The information in Indo-Iranian tradition about the Arctic and 
northern lights was borrowed by the Greeks from the Scythians (Bongard-Levin 
and Grantovsky 1983) and can be connected with the spread of Andronovo 
bronze axes and tools to the Arctic Circle. 

All this evidence supports the hypothesis that the Indo-Iranian homeland was 
situated in the neighborhood of the Finno-Ugrian zone and, therefore, one may 
assign the Indo-Iranians to the Andronovo culture. 

Indo-Iranians and Greeks 

Scholars have often stressed the special importance of areal isogloses and 
mythologemes between the Indo-Iranian and Greek worlds (Elizarenkova 1989: 
438; Ivanov 1969b; 1974; Ivanov and Toporov 1960: 15-16). As Mycenaean texts 
show that the Achaeans were in Greece already in the 16th century BC, the 
culture claiming Indo-Iranian attribution must archaeologically exhibit ties with 
the Mycenaeans. S. S. Berezanskaya (1971) was the first to pay attention to 
contacts between the population of the Pontic steppes and Mycenaen Greece in 
the Bronze Age, which is confirmed by the resemblance in the architecture of the 
stone -built burial chambers and by some categories of inventory. I (1977; 1978; 
1980a; 1980b; 1981a; 1984b) have already mentioned the similarity of some 
object types from Petrovka, early Andronovo and Early Timber-grave sites, 
including arrows (important because in the Greek language an Iranian term is 
used for the bow and quiver, cf Benveniste 1937). The main item of comparison 
is the spread into Greece of horses for chariotry, horse burials and cheek-pieces 
which compare well with those of Andronovo and show the specific importance 
of horse-drawn chariots in both regions and their common origin. The resem- 
blance between archaic steppe cheek-pieces with tenons and Greek ones has 
been examined in Chapter 8. This similarity is of great importance, and to this 
we might add the spread of Mycenaean ornaments on the cheek-pieces (Karo 
1930; Penner 1998), scepters and other objects in the steppes, especially along 
the Don (Figs. 38, 66, 67, 79). These facts are extremely important because 
myths connected with the horse, chariot and fast -running solar horse, are related 
to a number of regional Greek-Indo-Iranian correspondences (Vanderpool 1959; 
Crouwel 1980: 22; Nefedkin 2000: 111; Smirnov and Kuz'mina 1977: 42-50, 
fig. 1 1-12). At present new samples have been found on steppe sites proving the 
existence of connections with Greece (Sinyuk et al. 1985; Pryakhin et al. 1988a, 
b; Penner 1998). Current criticisms of these conclusions (Besedin 1996; Trifonov 
1996) are not convincing (Kuz'mina 1998a, b; Penner 1998). 

In Greek myth Apollo, the Sun god, was born in the north, in the Ripa 
mountains. In the vernal equinox white swans harnessed to a chariot carry him to 
Greece (Ivanov 1969: 54; 198). Ripa is the name of the Ural mountains, which 
are situated according to Indo-Iranian mythology in their homeland (Dovatur et 
al. 1982). In the Urals, on the vernal equinox, there is a mass swan migration 
(Chlenova 1983a). These facts support the existence of the areal Greek-Indo- 
Iranian contacts and the corresponding localization of the homeland in the 
steppes, to be more exact, in the Urals. 



CHAPTER FOURTEEN 
CONCLUSIONS TO PART ONE 

This analysis of the material culture of the Andronovo tribes has been conducted 
according to the methodology outlined above and over a vast background of 
synchronous and diachronous archaeological cultures. The evidence has been 
viewed in the context of written documents, and linguistic, anthropological, and 
ethnographic data on the culture of the Indo-Iranian peoples; its conclusions have 
been verified independently by a variety of methods: 1) by establishing the time 
and direction of migration routes; 2) the analysis of anthropological data; 3) the 
combining of archaeological and toponymic maps; 4) the study of Indo-Iranian 
traditions about their homeland; 5) the possibility of interpreting the semantics of 
Andronovo art and rituals in the light of Indo-Iranian mythology and ritual prac- 
tice; and 6) the evidence of contacts with Finno-Ugrian and Greek speakers. All 
of these have led me to the following conclusions. 

The retrospective method establishes the succession of the Andronovo 
culture and the cultures of the Iranian-speaking Sauromatians and Saka, which 
replaced it in the 8th century BC. This follows not from an individual element of 
the complex but by the sum total of features, both belonging to one ECT (which 
defines the similarity of tools, composition of herd, partially house type) and 
ethnically diagnostic features, which are not connected with the sphere of 
manufacture and which as a result give extremely important information about 
ethnic succession thus serving as ethnic indicators (ceramics, clothes, ornaments, 
types of industrial technology, range of prestige animals reconstructed through 
the study of ritual burials and art that reflect a unity of mythological ideas, and 
the retention of functionally unconditioned traditions in production). The genetic 
connection between specific sets of weapons (spears, arrows), and horse gear 
(cheek-pieces, vehicles) is significant. The genetic succession of the steppe 
population is supported by anthropological data. It helps affirm the conclusions 
of previous researches and accepts the Saka and Sauromatians as the direct 
descendents of the Andronovo tribes and thus renders the Iranian or Indo-Iranian 
attribution of Andronovo justified. 

Andronovo traditions are revealed in the culture of modern Iranian, Indian 
and relict tribes (domestic architecture in Ossetia, the Hindukush, the Pamirs and 
Hindustan, the type of light dwellings of Iranian nomads, the technique of hand- 
made pottery in the Pamirs and Hindukush, dress). The succession of 
anthropological type of some population groups has been traced. This too affirms 
the validity of attributing an Indo-Iranian identity to the Andronovo culture. 

The analysis of the entire Andronovo natural habitat, and the reconstruction 
of its ECT as an integral system has led us to assign it to the central Eurasian 
zone and marks its utmost proximity to the ECT of the pastoral cultures of the 
Eurasian steppes, especially the Timber-grave culture, and marks it as 



206 CHAPTER FOURTEEN 

fundamentally different from the Indo-Near Eastern ECT type, seen in its 
agriculture, form of stock-breeding and herd composition, industrial character, 
settlements and houses types, dress, etc. In the Andronovo and Timber-grave 
area of the 17th-9th centuries BC there is no mass migration and no major 
cultural influences from the second order civilizations of Iran and southern 
Central Asia. 

The combination method has shown that the ECT of the early Indo-Iranians, 
reconstructed by written texts, historic tradition, and linguistic and ethnographic 
materials, does not fit the Indo-Near Eastern zone but finds correspondences in 
the Central Eurasian zone in the circle of pastoral cultures of the Eurasian 
steppes. This is seen in the mixed economic type with stock-breeding dominant, 
herd composition (horse, cattle, sheep), character of industrial production 
(domestic production for the family, absence of specialized differentiated crafts 
for the market, hand-made pottery); house type (long term pole-supported 
structure with a ridged roof and light proto-yurt); dress which did not correspond 
to the ecological conditions of India and Iran and had no sources there, while the 
genesis of every category can be traced to the south Russian steppes from the 
Eneolithic period. The systemic character of connections is established not 
according to individual elements but according to a sum of interconnected 
culturally determining features which firmly connects the origin of the Indo- 
Iranian ethnos with cultures of the Eurasian steppes, first of all with Andronovo, 
Timber-grave and also with the related Tazabagyab and Vakhsh. They probably 
reflect the ethnographic cultures of different tribes of Indo-Iranians before their 
split in the steppe homeland long before they began their migration to India and 
Iran. Numerous Indo-Iranian tribes with cultural peculiarities and dialects are 
known from the Rigveda and Mahabharata. 

The method of surveying culture-indicative features specific only for the 
given ethnos was used as a further step in the comparison of Indo-Iranian 
culture, especially Indo-Aryan, with Andronovo. The most important ethnic 
indicators are: the absence of pig in domestic livestock; the presence of the 
Bactrian camel; the special role of horse-breeding and horse sacrifices; the 
special role of the horse-drawn chariot and its cult; the technique employed in 
making tripartite vessels by means of coil modeling; the form of unique square 
vessels; and cremation ritual. These confirm the validity of our attribution and 
make it possible to stress the specific similarity of some Fedorovo type features 
with Indo-Aryan. 

The process of ethnogenesis in the steppes in the 2nd millennium BC was of 
an autochthonous character of development, involving integration and migration, 
which was strengthened in the 17th— 16th centuries BC, probably because of the 
appearance of the chariot and bronze casting. In the west of Andronovo territory, 
the Sintashta-Petrovka type sites were formed as a result of the influence of wes- 
tern cultures (Poltavka, Abashevo, Catacomb, Multi-roller Ware); the Timber- 
grave culture was formed simultaneously under the influence of the same 
components. 

Beginning from the middle of the 2nd millennium BC there was intensive 
assimilation and integration in the Andronovo region of the Alakul' tribes, that 
are genetically connected with the Petrovka and Fedorovka tribes, which led to a 
consolidation of the Andronovo culture. In the third quarter of the 2nd 



CONCLUSIONS TO PART ONE 207 

millennium BC there was a settling apart of groups of the Andronovo population 
from Kazakhstan to south Siberia and Central Asia. Gradually the pastoralists 
mastered new ecological niches; the highlands of the Pamirs and Tian-Shan and 
the deserts of Central Asia. The first wave of Indo-Iranian migration to the south 
is dated to the Novokumak period. 

Population movements in the steppes were governed by the character of their 
pastoral economy that demanded a change of pastures every 20-25 years, and 
with increased livestock there was a need to widen the areas of pasture. This led 
to the rise of more progressive semi -nomadism and then to the nomadic form of 
pastoralism. The premises for mastering new territories were slowly developed 
in the 2nd millennium BC as a result of the normal development of the pastoral 
economy: a herd (cheese, koumiss); suitable dress; orientation skills; discovery 
of wells. The complex of these conditions created the possibility for migration to 
the south. Its reasons were concealed in the very character of extensive stock- 
breeding and the unstable ecology of the steppes. 

The penetration of the speakers of the various Indo-Iranian dialects to the 
south was a process which was both long and which occurred at different times. 
Also, it did not involve the mass change of the aboriginal population. This type 
of migration to a zone of a more progressive ECT involves craftsmen remaining 
aborigines and newcomers acquiring the cultural complex by adjusting to the 
new ecological niche. It is archaeologically established first of all by changes in 
the spiritual sphere and by separate innovations in material culture that serve as 
ethnic indicators. At the end of the 2nd millennium BC in the south of Central 
Asia, in Iran and in the north of Hindustan such innovations comprised the 
appearance of hand-made ceramics, technologically non-conditioned ornamen- 
tation with modeled rollers; a wooden house with ridged roof; proto-yurt; and 
costume. These artifacts had no sources in the local culture and were ecological- 
ly alien to it, as well as the horse, chariot, and camel cult reflected in art and 
ritual and tumulus burial and cremation. Their appearance in the south witnesses 
the arrival of a new population from the Asian steppes. It is possible to connect 
the spread of different Indo-Iranian dialects with this population. 



PART TWO 

THE MIGRATION OF TRIBES AND THEIR CULTURES IN 

CENTRAL ASIA 



CHAPTER FIFTEEN 

CULTURES OF CENTRAL ASIA IN THE 4TH-3RD MILLENNIA BC 

A number of economic and cultural zones formed in Central Asia during the 4th- 
3rd millennia BC. In the south, in Turkestan (the future Parthia), the lowland 
territory and the lower reaches of the Tedjen river saw the development of the 
farming culture of Anau brought about by migration from Iran (Pumpelly 1908), 
orNamazga (Masson 1966, 1982; Masson and Sarianidi 1972; Kohl 1981). 

According to the chronologies of B. Kuftin and V. Masson, stages I to III 
date to the Eneolithic. In the Namazga III period a new wave of migration from 
Elam into the oasis of Geoksyur is presumed which is documented by the simil- 
arity of ceramic ornament. The Eneolithic culture is represented by multi-layer 
settlements — tepes made up of blocks of multi-room houses and sanctuaries 
(Masson 1962, 1982: 58-60; Sarianidi 1962)— Anau (the northern hill), 
Namazga, Kara-depe, the Geoksyur oasis on the Tedjen. The economy was 
based on irrigation farming supplemented by the breeding of sheep, goats, 
Bactrian camels, and a southern breed of zebu-like cows (Tsalkin 1970; 
Ermolova 1976). Ceramics were hand-made and painted. Metal articles were 
made of imported forged copper (Kuz'mina 1966; Terekhova 1975). The dead 
were buried in the settlements under the house floors, and in the Namazga III 
epoch — in round burial-vaults or tholoi. The clay figurines reflect the cult of a 
mother-goddess. 

The Bronze Age, the stages of Namazga IV and V, is marked by the 
appearance of large fortified towns: Namazga at 50ha, Altyn-depe at 26ha, Ulug- 
depe at 20ha, and by the development of full craft specialization. The potter's 
wheel was employed for making pottery (Masimov 1976). Special furnaces were 
used for firing pottery and casting metalwork Four-wheeled wagons drawn by a 
pair of oxen or camels became common (Kuz'mina 1980b, 1983; Fig. 94). 

The Anau culture reached its florescence in the Namazga V period (Masson 
1964, 1966, 1976, 1981, 1982; Masson and Sarianidi 1972). Social stratification 
occurred. At the capital settlement of Altyn-depe one can distinguish blocks with 
large houses for the nobility and those of craftsmen engaged in making 
unornamented standardized ceramics and various articles of arsenical bronze, not 
forged but cast according to the lost-wax method. Rich burials came into being. 
Ideological concepts became more complicated, as evidenced by a four-tiered 
cultic construction at Altyn-depe, clay figurines of goddesses, and crosslike 
metal objects and zoomorphic seals. Imports from Tepe-Hissar and Harappa 
reflect wide cultural relations. 

Simultaneously with the development of the farming culture of Anau, the 
settlement of Sarazm emerged in the fertile valley of the Zeravshan, 45km east 
of Samarkand (Isakov 1991; Lyonnet 1996). It was founded by the former occu- 
pants of the Geoksyur oasis in the lower reaches of the Tedjen with a view to 



2 1 2 CHAPTER FIFTEEN 

mining the rich deposits of polymetallic ores and turquoise. These riches 
attracted also the farmers of Baluchistan, who constitute the second component 
of the Sarazm culture. The conclusions concerning the genesis of this culture are 
based on the analysis of the ornaments of the ceramic complexes and imported 
articles. In the development of Sarazm four stages are distinguished, dated by the 
radiocarbon method from 3400 to 2000 BC. Sarazm is an enormous settlement of 
lOOha, consisting of multi-room houses with temple -sanctuaries, and in Period 
III — with palaces. The economy was based on irrigation and, partially, bogar 
farming (i.e., semi-irrigated and only in spring) and the breeding of ovicaprids 
and cattle. In periods I-III the ceramics of Sarazm were hand-made and painted; 
in period IV (2700-2000 BC) the potter's wheel appeared. The working of 
turquoise, lapis lazuli, and agates was developed. Yet the staple branch of the 
economy throughout these periods was the extraction and smelting of metal, 
which is evidenced by the findings of furnaces, clay molds, crucibles, a lead 
ingot weighing 10kg as well as numerous metal articles. We thus may consider 
Sarazm as the largest metallurgical center of Central Asia engaged in export. 
Sarazm enjoyed wide relations from Iran up to Baluchistan and Seistan. In the 
3rd millennium BC wide contacts were characteristic of the ancient Near East as 
a whole (Amiet 1986; Lamberg-Karlovsky 1994). 

Another outpost of the ancient Near East civilization was the settlement of 
Shortughai in the north of Afghanistan in ancient Bactria, located on either bank 
of the Amu-Darya (Francfort 1977, 1989). The excavations of the hills of the 
settlement brought out the dynamics of its development. Four stages were 
distinguished. The settlement came into existence as a colony of the Harappan 
culture, which is documented by the typically Harappan building technique, the 
ceramics bearing Harappan decoration made on the potter's wheel, a seal, beads, 
and articles made of shell. The Indian colony is likely to have been set up in 
order to develop the rich neighboring deposits of lapis lazuli; by controlling this 
trade it was possible to establish wide cultural relations. Based on the 
radiocarbon dates H.-P. Francfort sets stage A to 2200-2000 BC and points out 
relations with Harappa, Mundigak IV, Shahr-i-Sokhta III-IV, Yahya IVB - 
Bampur V-VI, Kulli, Shahdad, and Hissar IIIC. 

Possibly, another center of the 3rd millennium BC farming culture will be 
discovered in the fertile valley of Fergana. Our hopes are aroused by the finding 
here of the Khak hoard containing copper articles of the Hissar III type and of 
two stone weights of ancient Near Eastern type (Voranets 1956; Kuz'mina 1966, 
Besenval 1987, fig. ii). 

Thus in the 3rd millennium BC the lands of southern Central Asia and 
northern Afghanistan, suitable for irrigation farming or rich in mineral resources, 
were occupied by a farming population that had migrated from various regions 
of the ancient Near East: from Iran (including Elam), Baluchistan and India. The 
migrants brought with them the achievements of Near Eastern civilization: the 
producing economy (domesticated wheat, barley, ovicaprids and cattle, including 
zebu-like cattle), the skills of manufacturing painted ceramics, metallurgy and 
metalworking as well as wheeled transport, which enabled wide exchange 
relations to be established. 

What is the ethnic origin of the various groups of farmers of southern Central 
Asia's south? 



CULTURES OF CENTRAL ASIA IN THE 4TH-3 RD MILLENNIA BC 213 

I. M. D'yakonov (1967: 85-87, 108-112) suggested that the Elamite language 
of south-western Iran and the Dravidian language of India were related and that 
they separated prior to the 3rd millennium BC. This hypothesis was substantiated 
by D. MacAlpin (1981). Dravidian is considered to have been the language of 
the population of southern Turkmenia (Litvinsky 1963: 128; Masson 1977: 115- 
118). H.-P. Francfort (1989) assumes that the Bactrians spoke Elamite. The 
majority of serious linguists hold that the Harappan civlization was created by 
Dravidians (Parpola 1974, 1988; Witzel 2001) or Dravidians and Munda (Bon- 
gard-Levin and Gurov 1988; 1990; Kuiper 1993). 

At the end of the 3rd millennium BC the ancient farming communities in the 
wide zone comprising the south of Central Asia, Iran, Afghanistan, Baluchistan 
and the south-east of Hindustan underwent a crisis. There is evidence of popula- 
tion decline (Namazga fell to 2ha), other settlements generally fell into a decline, 
e.g., Altyn-Depe, Sarazm, Shortughai, Tepe-Hissar III, Shah-Tepe, Tureng-Tepe, 
Shahr-i-Sokhta, Tepe-Yahya, Mundigak; and the Harappan civilization in India 
perished. What caused the crisis is not yet clear. Scholars have proposed 
ecological disasters: earthquakes, floods, drought, and soil exhaustion. M. 
Wheeler (1968) attributed the fall of Harappa to an Aryan invasion from the 
north. However, the picture appears to have been much more complex. The 
florescence of culture in Central Asia led to a demographic explosion. The ecolo- 
gical resources of the oases became insufficient for feeding the increased popula- 
tion. At some settlements it appears as if the crisis progressed gradually. The 
most plausible hypothesis seems to be that which was advanced independently 
by B. and R. Allchin (1988) and G. Bongard-Levin (1963, 1979, 1981; Bongard- 
Levin and Gurov 1988, 1990). In Eurasia during the late 3rd millennium BC 
there occurred an ecological crisis that aggravated the existing acute economic 
problems, thus triggering an internal political crisis and gradually leading to the 
widescale disaster. The collapse of most of the large settlements enabled the 
pastoral tribes that were situated on the borders of the farming oases to occupy 
the partially abandoned lands and to subjugate and gradually assimilate the 
indigenous population. As a result, following the crisis period, there formed new 
cultures. However, let us remember that this is just a hypothesis. 

If in the 4th-3rd millennia BC the oases of south Central Asia continued the 
farming cultures of Western Asia and India, what was happening in the same 
period in the other regions of Central Asia? 

In Tadzhikistan there existed an extremely primitive culture, the Hissar 
(Ranov 1998). This population was engaged in hunting and used archaic stone 
implements. In the north over the vast territory embracing the northern shores of 
the Caspian Sea, the region near the Aral Sea (future Chorasmia/Khorezm) and 
the Buchara oasis in Transoxiana there emerged the Kelteminar culture 
(Vinogradov 1968, 1981; Vinogradov et al. 1996; Gulyamov et al. 1966). The 
specific ecological conditions predetermined the main economic and cultural 
development of this region. Paleoecological research has established that during 
the 4th-3rd millennia BC this was the period of the Lyavlyakan pluvial 
(Vinogradov and Mamedov 1975: 234-55), when the present-day deserts were a 
series of impassable marshes and lakes. The lakes abounded in fish and water- 
fowl that made fishing the staple of the Kelteminar economy. The population 
dwelt in camps in large clan cabins (shelters). Though over 800 Kelteminar sites 



2 1 4 CHAPTER FIFTEEN 

have been examined, none have yielded either cereals or the bones of 
domesticated animals. "There are no elements of the producing economy... 
present" (Vinogradov 1981: 146, 147). 

A. A. Formozov (1959: 155) drew the boundary between the zones of Euro- 
pean and Asiatic Neolithic cultures along the Emba river. Subsequent research 
has confirmed the cardinal differences in the development of the Neolithic and 
Eneolithic cultures in the European and Asian zones of the steppes, which 
enables one "to speak of their belonging to different cultural domains" and to 
assign the sites of the Volga Region, the Urals and the north Caspian to the 
western Pontic-Caspian circle (Ivanov and Vasil'ev 1995: 121). In this zone in 
the 4th-3rd millennia BC one can trace a gradual development of the producing 
economy under the influence of the farming cultures of the Carpathians and the 
Danube (Vasil'ev and Vybornov 1988; Vybornov and Sinuk 1985; Rasamakin 
1999; Kuz'mina 1996-97; 2003). 

By contrast, in the Asian part the proximity of the southern farmers did not 
result in the formation of an agricultural economy and the subsistence of 
Khorezm's Kelteminar population was based on fishing and hunting water-fowl, 
and in the steppes of Kazakhstan — on the seasonal hunting of ungulates (for a 
discussion about an independent center of horse domestication in Kazakhstan see 
PSAH). These ecological and archaeological data rule out the very possibility of 
a mass migration of Indo-European farming populations from the Middle East 
through Central Asia into the Pontic-Caspian steppes, which was suggested by T. 
Gamkrelidze and V. Ivanov (1984) and all the more so of the localization of the 
Indo-European homeland in Central Asia, as suggested by J. Nichols (1997), or 
in Bactria (Sergent 1997). On the contrary, not only was there no migration from 
Central Asia into the steppes but also the contacts of the farming population with 
the northern tribes were very sporadic and did not lead to a change in the 
economic and cultural type. In Sarazm unprepossessing ceramics were found 
comparable with those of the Kelteminar and Pit-grave-Poltavka cultures 
(Lyonnet 1996: 49, pi. v, vi: 8). A. Vinogradov (1968: 146-150, fig. 62) 
identified some influence of the southern forms and ornaments on Kelteminar 
pottery in Khorezm, but it was only on the late Kelteminar settlement of 
Lyavlyakan 506 that molds for the ax-adze and ax-hammer of the southern types 
were found (Vinogradov and Kuz'mina 1970) along with a workshop for 
preparing turquoise; and in the Dam-Dam Chashme and Jebel caves in Trans- 
Caspia bones of sheep were discovered, probably domesticated, and ceramics of 
the Shah-Tepe III-II and Kyzyl-Arvat type (Tsalkin 1956: 220, 221; Shnirel'man 
1980: 74, 75). But these contacts were sporadic. 

On the other hand, during the drought in the steppes and the completion of 
the Lyavlyakan pluvial there can be traced an eastward advance from the 
territory of the Pontic-Caspian steppes of late Pit-grave tribes, apparently, under 
the influence of the Catacomb culture. They brought to Siberia farming and 
cattle-breeding skills, cereals and domesticated animals as well as metallurgy 
and wheeled transport and established the Afanas'evo culture (Vadetskaya 1986; 
Tsyb 1984; Semenov 1993; Gryaznov 1999: 51-54). Many associate the 
Afanas'evo culture with the Tocharians (Pulleyblank 1966, 1996; Mair 1998; 
Mallory and Mair 2000). 



CULTURES OF CENTRAL ASIA IN THE 4TH-3 RD MILLENNIA BC 215 

Some similarity to the sites of the Pit-grave culture are seen in burials from 
the cemeteries of Ubagan and Verkhnyaya Alabuga on the Tobol river 
(Potemkina 1985: 151-155, 201; fig. 63: 1-9) and in a mound surrounded by a 
stone circle from Karagash in Central Kazakhstan (Evdokimov and Loman 1989: 
34-46; fig. 5). The dead were buried in keeping with the Pit-grave rite on their 
backs with the knees raised, accompanied with ocher, copper and stone articles 
and round-based ceramics with stamped ornament. In Karagash the bones of 
sheep and cattle were found. Anthropologically, this burial is of the Pit-grave 
type. 

Under the influence of a western impulse at the end of the 3rd millennium 
BC we find for the first time among the late Neolithic sites of Kazakhstan the 
bones of domestic ovicaprids and cattle of East European breeds and also those 
of the horse (Formozov 1950; Chalaya 1977; HKSSR 1977; Makarova and 
Nurumov 1989). 

The earliest culture in Central Asia reflecting the active ties between the 
farming and steppe populations is the Zaman-Baba in the oasis of Bukhara 
(Gulyamov et al. 1966: 118-186; Kuz'mina 1958, 1968). The steppe component 
is seen in the form of a semi-subterranean hut at the settlement and the burial rite 
of the Pit-grave and Catacomb type in a kurgan, the presence of chalk, ocher, 
animal bones, and a form of round-based and pointed-based ceramics 
comparable with the Pit-grave and in part Catacomb types, yet devoid of 
ornament. Having arrived at the Zeravshan river, the pastoral population came 
into contact with the farmers, representatives of the late Sarazm and Namazga V 
cultures of the Hapus-depe settlement. 

The southern component is expressly represented by wheel-made ceramics, 
accounting for 10% of the vessels, ceramic kilns, painted vessels of the 
Mundigak IV type, two female figurines and copper pins with a shovel-like top, 
mirrors, various beads of turquoise, lapis lazuli and other stones as well as beads 
of paste made in the shape of a Maltese cross. 

The economy of the Zaman-Baba culture was complex: imprints of wheat 
and barley grains were found along with grinding stones and sickle blades; in the 
osteological assemblage wild animal bones constituted 15%, while 85% 
belonged to the cow, sheep and donkey, domesticated in the south. 

The cultural characteristics of the Zaman-Baba culture do not make it 
possible to assign it to the BMAC and date it to the second quarter of the 2nd 
millennium BC (Askarov 1981) or to the turn of the 2nd-lst millennium BC 
(Sarianidi 1979). A wide range of analogies with Namazga V, Shah-Tepe II, 
Hissar IIIB, and Mundigak IV determines its date as the first quarter of the 2nd 
millennium BC or earlier if the radiocarbon dates are to be accepted. The cultural 
synthesis, for the first time attested in Zaman-Baba, then becomes characteristic 
of the cultures of Central Asia throughout the historic epoch. 



CHAPTER SIXTEEN 

THE INDO-ARYAN MIGRATION AND THE FIRST STAGE OF THE 
ANDRONOVO MIGRATION TO THE SOUTH 

The problem of migration 

To verify the hypothesis that the Indo-Iranian homeland lay in the Asiatic steppe 
the issue of migrations both within the Eurasian steppes and neighboring regions 
is very important. First we need to examine the problem of recognizing 
migrations in the archaeological record. 

Gordon Childe's The Aryans: A Study of Indo-European Origins was 
published in 1926. Here he examined the problem of the Indo-European 
homeland. This fruitful trend was however for a long period compromised by the 
school of Gustav Kossina whose works were strongly nationalistic. During the 
difficult years of the ideological struggle of the 1930s, Soviet archaeologists 
transferred N. Ya. Marr's stadial theory, propagated initially in linguistics, to 
archaeology. They rejected the idea of migration and began to support an auto- 
chthonous theory. A. A. Formozov (1959b) must be honored for acknowledging 
the role of migration as an important factor in ethnogenesis. Migration is a com- 
plex historic process. N. Ya Merpert (1978: 9-27) singled out types of migration 
of the Neolithic, Eneolithic and Bronze Age. D. Dietz (1973: 1-14) worked out a 
template of migration criteria with a complete transfer of culture. R. Hachmann 
(1970) and L. S. Klein (1973: 1-14) noted cultural transformation during the 
adaptive process to a new ecological niche and the inclusion of new groups in 
the process of migration (corporative by-pass migration) and spoke skeptically 
about the possibility of identifying them. V. S. Titov (1982: 89-1 14) paid special 
attention to the reflections of anthropological migration and the external signs of 
invasion (fires, hoards, etc) and he developed a tree models of substratum - 
superstratum relations. C. Renfrew (1987: 120-144) stressed the role of social 
and economic factors and outlined three models: demographic (expansion in 
search of resources, often gradually from one center); elite dominance (sub- 
mission of aborigines by a small group of emigrants); and the model of systems 
collapse (economic and social crisis of aboriginal culture simplifying migration 
from one center or from some periphery). This model had already been 
employed by G. M. Bongard -Levin (1963) in connection with the arrival of the 
Indo-Aryans in the Harappan civilization. Yu. V. Bromley (1983: 273-283) em- 
phasized the role of migration in ethnogenesis and the process of social stratif- 
ication and the importance of mixing and bilingualism. V. P. Alekseev (1986: 5, 
8-10) discussed the possibility of preserving an aboriginal anthropological type 
during an ethnic shift, focusing on the characteristics of the newcomers during 
intermarriage and bilingualism. I. M. D'yakonov (1982) approached the problem 
on the basis of documentary analysis and demonstrated that migrations in the 



2 1 8 CHAPTER SIXTEEN 

Near East, even when they led to language shift, were not always reflected in the 
archaeological record. I (Kuz'mina 1981a: 103, 104; 1986c: 174, 175) studied 
the question of the reflection of migration in the economic and cultural zones 
(ECT, cf. p. 13): 

A. Migration from a more advanced ECT to the zone of a less advanced type. 

1. Colonization (or systematic migration): the culture and language of the 
superstratum are completely transferred; the substratum population is ousted or it 
survives within its own ECT. The archaeological reflection is a widening of the 
culture's natural habitat (including discrete settlements). Examples are the col- 
onies of the Harappans, Phoenicians, Greeks, Roman camps, Russian settlements 
in Siberia. This type of migration must not be confused with cultural borrowing. 

B. Migration within a single ECT. 

1. Colonization (or systematic migration): the culture and language of the 
superstratum are completely transferred; the substratum population is ousted. 
The archaeological reflection is an expansion of the culture's natural habitat. 

2. Assimilation: migration in one direction (single trajectory) or in different 
directions from one center (radial), often slow and in stages, resulting in steady 
contacts with the substratum population that influences the culture greatly. The 
archaeological reflection is the formation of a local culture variant, and with 
continuous isolation we expect divergence. The superstratum language shares 
isoglosses with the substratum; sometimes, if bilingualism continues the substra- 
tum language wins. Conclusions about language shift or preservation are 
impossible without additional data. 

3. Integration: the migrating bearers of one culture (single trajectory) or an 
amalgamation consisting of the bearers of several cultures, leads to the 
interaction between the substratum and one or several superstratum cultures. The 
archaeological reflection of this scenario would be the formation of a new 
archaeological culture as a result of the combination of its several elements 
('blocks' to use the language of V. S. Bochkarev). There is a substratum 
language or one of the superstrata with isoglosses of others. Conclusions 
regarding the language are impossible without additional data. 

C. Migration from a more backward ECT to a more advanced. 

1 . Colonization which in this zone presupposes a mass invasion (in contrast to A 
and B) as a result of a conquest which is often attributed to a preceding crisis or 
collapse of the substratum culture due to internal social and economic reasons. 
The archaeological reflection sometimes consists of traces of violent destruction, 
fires, hoarding on sites of the substratum culture, which are covered by a layer of 
a new culture; there may be the formation of a new regressive culture, in which 
gradually the old superstratum elements re-emerge at the expense of the migrants 
who have had to adapt to a new ecological niche; there may be the inclusion of 
aboriginal craftsmen. Innovations appear in the spiritual culture (burial rite, art 
images), also in military art and the appearance of some features of culture and 
production that are not ecologically and functionally preconditioned. The 
superstratum language is established. In the case of an amalgamated migration 



THE INDO-ARYAN MIGRATION 2 1 9 

the substratum language can be preserved. Examples are Central Asia, the 
Turkish conquests of the Byzantine empire, the Barbarian conquest of the 
Roman empire. 

2. Elite assimilation and integration: the migration of a small group, unified and 
possessing a military advantage and establishing its political supremacy. Here 
the newcomers adopt the aboriginal culture. The archaeological reflection is the 
preservation of the substratum culture, innovations in spiritual culture (burial 
rite, iconography) in some ecologically and functionally unconditioned elements, 
sometimes in military technology, defensive architecture, aristocratic burials that 
point to social stratification. It is impossible to draw conclusions on the preserva- 
tion or shift of language without additional data. In the case of a superstratum 
victory (integration) it fixes borrowings from the cultural and social vocabulary 
(scribe, literary texts, priest, temple, palace, specialized craftsmen), notations of 
a new type of economy and a new ecological sphere (including semantic shift, 
e.g., Proto-Indo-European 'willow' in Hittite becomes 'olive-tree'). Deities of 
the substratum religion are incorporated. In the case of a victory of the 
substratum language (assimilation) this establishes onomastic innovations, some 
lexical borrowings, especially in military technology. A new language is formed 
in the case of long bilingualism. 

Thus, the reflection of migration in culture, and hence in archaeological 
culture, is in the first place dependant on the ECT of the migrating participants 
and its direction; the direction of language shift is even more difficult and it is 
probably dictated largely by social factors. Anthropological traces of migration 
are sufficiently clearly fixed in a change of anthropological type in the case of a 
mass migration of a population, that replaces the aboriginals or in the appearance 
of mixed types in the case of intermarriage and processes of integration; here the 
substratum type often dominates. These three processes: culture, language and 
change in anthropological type are sometimes independent and differently 
directed, and ethno-genetic conclusions are possible only as a result of a complex 
comparison and mutual correction of all kinds of sources. 

Analysis based on a large body of material concerning migrations, drawn 
from written sources, permits one to posit possible models of Indo-Iranian 
migration. According to the hypothesis of V. V. Ivanov and T. V. Gamkrelidze 
we should expect a migration of type 1, i.e., the colonization of the steppe by 
Iranian farmers, who brought with them from the Near East the whole complex 
of civilization, and into eastern Iran and India they carried an archaeological 
culture formed in western Iran. Both the first and the second supposition are 
archaeologically disproved: even supporters of the early appearance of the Indo- 
Iranians in Iran and their association with Grey Ware (T. C. Young, J. Deshayes, 
R. Dyson, I. N. Khlopin) posit its origin in the 3rd millennium BC in eastern Iran 
and a later spread to the west only at the end of the 2nd millennium BC from the 
east. V. I. Sarianidi (1990) localized the Indo-Iranian complex in eastern Iran and 
argued its spread into Baluchistan and to the south of Central Asia. At present he 
suggests connections of the Proto-Indo-Iranians with Greece (without taking into 
account the chronology of Achaean migrations) and the arrival of Indo-Aryans 
from the Near East; he connects their culture with the Bactria-Margiana 
Archaeological Complex (BMAC). In the latter he sees the main sources of 
Zoroastrianism (Sarianidi 1998). There is serious criticism of this hypothesis 



220 CHAPTER SIXTEEN 

(Mallory 1988; VDI 1989). Neither of these hypotheses explains the formation 
of the Indo-Iranian steppe peoples: the Scythians, Sarmatians and Saka, as well 
as the early contacts between the Indo-Iranians and the Finno-Ugrians. 

This does not permit an acceptance of the Near Eastern hypothesis and it 
confirms a steppe localization as the Indo-Iranian homeland. Moreover, abundant 
evidence for the cattle-raising character of Indo-Iranian economy in their 
language and in the Avesta and Vedic tradition are totally neglected. 



Migration in the Eurasian steppe and neighboring regions 

Until the present there have been two competing models for the Bronze Age of 
the Eurasian steppe: migration and autochthonous development. The main 
achievement of the past decades has been the establishment of a new view of the 
process of cultural formation. In the history of the steppe there are periods that 
are probably connected with cultural innovations, e.g., the establishment of 
settled stock-breeding and its replacement by pastoral nomadism, the transition 
to bronze or iron metallurgy, the mastery of chariots or horse-riding. When the 
smooth cultural development of the steppe is seen to be broken we find 
considerable evidence for the migration of ethnic groups and the regrouping of 
cultural features. As a result, new cultures are formed on the basis of similar 
components and these are termed 'cultural blocks'. Their specific composition is 
conditioned by the varying degrees of substrate elements and the interaction of 
cultures that come into contact during corporate migration. 

Uncovering the direction of steppe migrations during the 2nd millennium BC 
is the most important method employed in verifying the hypothesis that the 
Andronovo tribes were Indo-Iranian. What were the ethnic movements and what 
was the nature of interactions across ethnic units? 

We can discern three periods in the history of the Eurasian steppe: 1) the first 
quarter of the 2nd millennium BC when earlier cultures decayed and there 
emerged two large cultural entities — the Timber-grave culture of the European 
steppe and forest-steppe and the Andronovo culture of the Asian region; 2) the 
15th 13th centuries BC when there was a period of stabilization and the Timber- 
grave and Andronovo cultures engaged in extending their territories through the 
colonization of new lands and by forcing out or suppressing native populations 
whose influence would precondition the formation of local variants; and 3) the 
12th— 9th centuries BC when there was considerable increase in activity between 
cultures and migrations primarily to the south and east. 



The first stage of Andronovo migration 

At the end of the 3rd/first quarter of the 2nd millennium BC there were 
considerable ethnic movements across the Eurasian steppe. These were probably 
stimulated both by an ecological crisis — increased steppe aridity (Kremenetsky 
2000; Spiridonova 1991) and the consequential introduction of pastoral 
nomadism — and the improvement of both transport and bronze metallurgy (Map 
8). The Poltavka culture formed on the basis of the Pit-grave expansion from the 



THE INDO-ARYAN MIGRATION 22 1 

Middle and Lower Volga to the southern Urals and the north Caspian region as 
far as Mangyshlak (Fedorova-Davydova 1964: 23-90; Kuz'mina 1976: 25, 26; 
materials from site surveys by I, B. Vasil'ev, A. N. Melent'ev, V. D. Beletsky; 
Vasil'ev et al. 1986; Problems of Ancient History 1990). The Don saw the 
migration of Catacomb tribes (Kiyashko 1974: 18; Bratchenko 1976: 117-118; 
Kachalova 1972). Catacomb and then Multi-roller sites appeared along the 
Volga. The Abashevo culture moved to the east (Kachalova 1976a; Kachalova 
and Vasil'ev 1989; Kuz'mina, O. 1979; Malov 1986; 1989). 

G. B. Zdanovich (1979a; b; 1988) surveyed the early Andronovo sites of the 
Petrovka type in northern Kazakhstan and he suggested that they were formed on 
the basis of the local Neolithic Vishnevka type complex although he also 
admitted some influence from the western Abashevo culture and possible ties 
with Central Asia. 

The discovery of the Sintashta cemetery in the Urals with its burials of 
warrior-charioteers accompanied by a rich set of arms, chariots, pairs of draught 
horses and cheek-pieces for their harnessing and burial 25 in the Novy Kumak 
cemetery prompted K F. Smirnov and myself (1976; 1977) to propose a Novo- 
kumak chronological horizon that unites sites of the Urals and western and 
northern Kazakhstan. On the basis of the stratigraphic position of the Novo- 
kumak kurgan the horizon can be set between the Catacomb culture and the 
developed Andronovo culture of the Alakul' type. At this time the characteristic 
cheek-pieces are found as far as Mycenae and the horizon is dated to the 17th— 
16th centuries BC. Most important is the bold suggestion that the Novokumak- 
Sintashta horizon formed in the Urals as a result of a migration from the west 
and the assimilation of several East European cultures — the Catacomb culture (in 
its later development as the Multi-roller Ware culture), the Abashevo culture 
(within the large number of Corded Ware cultures of Europe), and the Poltavka 
culture which reflects a later derivative of the Pit-grave culture along the Volga 
and in the Urals. 

The most important characteristics of the Novokumak sites that help define 
an ethnic group — kurgan burial, timber structure in burial chamber, ritual animal 
burials, cult of the horse and vehicle — derive from Eastern Europe; they are not 
found in the local Neolithic which itself may be traced to west of the Urals 
during the 4th-3rd millennium BC. The similarity among jars and pot-like 
vessels with a ridge on the shoulder among the Sintashta, Petrovka and Poltavka 
cultures is a product of their method of manufacture. Another important 
component of the Novokumak horizon is the Abashevo culture which is 
demonstrated by vessels of particular Abashevo form with projection on the rim 
of the vessel, Abashevo types of decoration and some features of the burial rite 
(flat graves). The similarity with the Multi-roller Ware culture is seen in the 
distribution within the Novokumak complexes of vessels that imitate Multi-roller 
types in form and ornament, modeled cones, herring-bone motifs divided by 
vertical lines (Smirnov and Kuz'mina 1977: fig. 9). V. F. Gening (1977) 
accepted the participation of western cultures in the formation of Sintashta. 

Potapovka burials and Utevka VI, which has been assigned to the Potapovka 
type, are related to the burials from the Urals; at present they are known in the 
Middle Volga region. They are considered to be the genetic predecessors of the 
Timber-grave burials (Vasil'ev et al. 1991; 1992; 1994; 1995). Elite warrior 



222 CHAPTER SIXTEEN 

burials with weapons and cheek-pieces that have been recovered from the Don 
region are assigned to the Don-Volga Abashevo culture (Pryakhin and Matveev 
1991; Pryakhin, Besedin et al. 1989; 1990; 1991; 1998; Pryakhin and Besedin 
1998; Matveev et al. 1995) or to the late Pokrovskiy-Abashevo culture (Sinyuk 
1996; Sinyuk e? a/. 1995). 

Stratified sites provide information concerning the formation and develop- 
ment of these complexes. The earliest Catacomb culture on the Don is replaced 
by the Abashevo culture which is, in turn, covered by the Timber-grave culture 
(Matveev et al. 1995: 78; Sinyuk et al. 1995: 40-49; Moiseev et al. 1995: 75). 

A series of Catacomb sites have been found on the Lower Volga which 
precede Timber-grave sites (Malov and Filipchenko 1995; Dremov 1996). The 
coexistence of the Poltavka and Abashevo complexes has been established for 
the Middle Volga in Aleksandrovka (Vasil'ev et al. 1995). At the settlement site 
of Kuysak in the Urals the lower layer exhibits Poltavka ceramics; the middle 
layer is mixed with Abashevo ware; and in the upper layer vessels of the 
Sintashta type are found which reflect a synthesis of the original components 
(Malyutina et al. 1995). Along the Ilek, isolated Poltavka burials are found in the 
cemeteries of Zhaman-Kargala, Tanabergen and Imangazy-Karasu; subsequent 
burials, situated in a circle and destroying the earlier ones, were of the Sintashta 
type (Tkachev 1996: 64; 1998: 42). At the cemetery of Bol'shekaragan in 
kurgans 1 1 and 24 one finds Poltavka and Sintashta vessels coexisting; the later 
ceramics of kurgans 20 and 22 are similar to those of Petrovka and the Timber- 
grave culture (Batalov et al. 1996: 86-88). 

There is evidence that permits us to distinguish two stages within the 
Novokumak horizon. On the settlement of Ust'e the earlier occupation belongs 
to the Sintashta type with its oval plan while this is covered by a later settlement 
of rectangular form which may be assigned to the Petrovka type (Vinogradov 
1995a: 17; 1999). The same sequence is evident from other settlements on the 
basis of aerial survey (Batanina 1995; Zdanovich G. and D. 1995: 50; Zdanovich 
1997: 59; Nelin 1999: 17). A similar development from Sintashta to Petrovka 
type has been observed in the cemeteries of Krivoe Ozero, Stepnoe, Kamenny 
Ambar, Bol'shekaragan, Tanabergen (Vinogradov 1995a: 26; 1999: 66; Kostyu- 
kov et al. 1995: 173; Batalov et al. 1996: 86-88; Tkachev 1995; 1998: 42-46; 
Epimakhov 1998: 180; Figs. 59-64). 

Thus, the formation of the complexes of the Novokumak horizon occurred 
due to the assimilation of tribes living in the vicinity of the Urals who belonged 
to the wide range of European cultures — Pit-grave, Poltavka, Catacomb and the 
more northerly Abashevo. They played different roles in the various regions 
which accounts for their local characteristics. They probably witnessed a long 
process of consolidation which presupposes a bilingual population. This is why it 
does not seem accurate to provide a purely Aryan attribution to the Sintashta 
people. 

There is no evidence so far for the participation of a local substrate in these 
processes (Matveev 1998) which contradicts the opinion of G. B. Zdanovich 
(1997). In the next stage, the Petrovka phase, there was a consolidation of 
culture, the bearers of which spread across northern and central Kazakhstan and 
into Central Asia; this reflects the beginning of the early migration of the 
Andronovo population. 



THE INDO-ARYAN MIGRATION 223 

Sites of the Novokumak horizon share similar features: the syncretic 
character of the culture and the appearance of chariots and the armament 
associated with the chariot -driving warriors. What was the reason for these 
innovations and what was the source? 

K. F. Smirnov and myself (1977) suggested that the impulse of these ethnic 
and cultural innovations was a migration by representatives of several cultures to 
the east. These cultures already formed the basis for future innovations which 
was stimulated by their contacts with more developed cultures along the Danube 
and the Caucasus. V. S. Bochkarev (1991) suggested the idea of a center of 
cultural genesis in the Volga-Ural region. This idea is shared by A. T. Sinyuk 
(1996) who believes that the new culture came to the Don from the Urals but 
other Voronezh archaeologists oppose this point of view and stress the absence 
of many features in the Urals which we find in the west in the Catacomb and 
Abashevo cultures. 

An ecological crisis (sharp aridization of the climate) could be one of the 
reasons for migration to the east (Kremenetsky 2000; Spiridonova 1991). 
Another and perhaps the most likely stimulus for migration to the Urals was a 
crisis in the Carpathian metallurgical center (Chernykh 1978) that used to supply 
metal to the steppe; now local exploitation of copper deposits was required. It is 
likely that by the late Pit-grave period excavation had already begun in the 
largest of the Kargaly deposits in the southern Urals (Morgunova et ah 1994; 
Chernykh 1999). Some small but easily accessible and rich copper mines were 
discovered near the Sintashta settlements (Zaykov 1995). Metal processing was 
the main occupation on all the settlements and in evidence in every house 
(Grigoriev 1994). It was the shift to the Urals as the major metallurgical 
production center of the Eurasian steppes that made the Urals (and not the Volga 
or Don) the center of cultural genesis. 

Traces of numerous fires and rebuilding have been found in all the Sintashta 
fortresses that have been examined; all central warrior graves were robbed in 
antiquity reflecting the unstable situation of the region. The necessity to defend 
the mines and the settlements of metalworkers required the building of fortresses 
and these then served as the production centers (Fig. 58). The labor requirements 
of building the fortresses, undertaking mining and the processing of the metals 
demanded a social group that could organize the workforce and distribute metal. 
The same group probably also oversaw the military functions. The main task of 
the organizing elite was the consolidation of the ethnic group. This military elite 
stimulated the developments of the arts of war and the introduction of the tactics 
of chariot warfare. The development of metallurgy and the use of the chariot 
expanded the cultural connections of the Ural population. 



CHAPTER SEVENTEEN 

THE AGRICULTURAL TRIBES OF SOUTH CENTRAL ASIA IN 
THE 2ND MILLENNIUM BC 

The culture of Central Asia is still divided into the south and north. In southern 
Turkmenia the 2nd millennium BC belongs to the Namazga VI stage (Pumpelly 
1908; Ganyalin 1956; Marushchenko 1959; Masson 1959; Shchetenko 1999). 
The date of this stage ranges from the (20th) 18th to the 13th centuries BC. The 
population failed to recover from the earlier crisis and the number of settlements 
decreased, e.g., Anau (Southern hill), Namazga-depe, Tekkem-depe, El'ken- 
depe. Their area does not exceed 1 -2ha. But the prolonged development of the 
stage is reflected in the thickness of the cultural layers, e.g., Namazga — 7m, 
Tekkem — 6 m. 

The culture is genetically related to the preceding Namazga V culture. Its 
traditions are preserved in the architecture of the fortified settlements with blocks 
of multi-room houses and in the manufacture of unornamented wheel-made 
ceramics, fired in two-tiered kilns, in the types of decoration and in metal 
objects. 

The economy is based on irrigation farming (barley and wheat). The 
importance of cattle-breeding increased: cattle, ovicaprids and Bactrian camels. 
The horse appeared for the first time. V. Masson (1959: 109, 1 10) notes progress 
in the development of craft and irrigation. Innovations are also recorded in the 
sphere of spiritual culture: special burial grounds (Yangi-Kala and Asgabat) 
came into being. Relations with India continued, indicated by the discovery of 
stone beads of the Jhukar type with circular ornament. Yet the orientation of the 
main contacts changes: ties are established with the northern pastoral tribes, and 
high-tin bronze articles of the Andronovo-type become widespread. 

In the 2nd millennium BC the center of the farming culture shifts eastwards. 
As early as the Namazga V and, possibly, even the Namazga IV period, a new 
seat of the ancient Near Eastern civilization forms in the delta of the Murgab 
river (future Margiana) and in northern Afganistan and south-east of Central 
Asia downstream of the Oxus river (Amu-Darya): the Oxus culture or the 
BMAC (Bactria-Margiana Archaeological Complex) (Sarianidi 1977, 1990; 
2002; Masimov 1979; Francfort 1989; Hiebert et al. 1992; Lamberg-Karlovsky 
1994). Both these names are inexact, since the Oxus culture characterizes only 
Bactria, while the BMAC is merely a complex, although we are dealing here 
with a full culture with several variants. 

The development of this culture dates back to the late (?) 3rd-2nd 
millennium BC. In Margiana the BMAC is represented by the major settlement 
of Gonur and its affiliated cemetery as well as the settlements of Kelleli, 
Togolok 1, 21, 24, Auchin I, and Takhirbay 3. In southern Bactria BMAC 
settlements include Girdai-Tepe, Dashly 1,3, Shortughai (stage B), numerous 



226 CHAPTER SEVENTEEN 

cemeteries destroyed by looters, the Fullol hoard, and chance finds (Sarianidi 
1977, 1990, 1998, 2001a,b, 2002; Francfort 1989; Amiet 1977, 1986, 1997; 
Pottier 1984; Tosi and Wardak 1972). 

The origin of the culture is debatable. H.-P. Francfort emphasizes the 
continuity of life at Shortughai and the continuity of building techniques during 
stage B of the Harappan cultural traditions. V. Sarianidi (1987) presumes that the 
culture emerged in eastern Iran where it has not yet been discovered. Now he, 
following T. Gamkrelidze and V. Ivanov (1984), points out its western sources, 
noting ties with Anatolia and even Greece. 

Initially, V. Masson (1989: 172,173) supposed that the development of the 
Murgab was associated with the cultural crisis of southern Turkmenia in the 
Namazga V period, when part of the excess population moved to the Murgab. 
This view seems unquestionable. The decisive argument in its favor is the 
continuity in the BMAC of the Namazga IV and V traditions of the manufacture 
of light-slipped ceramics, and the industrial technology that was maintained in 
Central Asia up to the time of Alexander the Great (Kuz'mina 1971, 1976). 

The ceramic complex of Central Asia during the Bronze Age and 
Achaemenid period differs from the gray-black polished pottery of north-eastern 
Iran (Shah-Tepe, Hissar), just as from the cultures of Baluchistan and India. 

The second component of the BMAC is the culture of Elam, as has been 
convincingly demonstrated by P. Amiet (1986, 1997). Of critical importance is 
not just the similarity of random types of articles but of the whole complex of the 
material culture (specific axes with a figured top, mirrors with an 
anthropomorphic handle, pins, steatite columns and female figures in Elamite 
dress) and, most importantly, images of art reflecting the spiritual culture of 
society. As a rule, their borrowing is associated not with cultural influences but 
with the migration of a population bringing along a new ideology. This 
conclusion is reliably supported by analysis of the semantics of the images on the 
seals (Klochkov 1997; Antonova 2000; Francfort 2001). As for ties with 
Anatolia, identified by Sarianidi, they are likely to reflect not a migration but 
wide cultural relations characteristic of the Bronze Age of the whole of south- 
west Asia (Amiet 1986; Lamberg-Karlovsky 1987, 1994; Hiebert and Lamberg- 
Karlovsky 1992). 

In the evolution of the Margiana culture several stages are singled out, 
reflecting the development of the oases on the Murgab. The earliest was the 
Kelleli oasis (Masimov 1979). At the settlement of Kelleli the Namazga V 
traditions are represented particularly clearly. The heyday of the culture goes 
back to the Gonur period. V. Sarianidi believes that a shift of the Murgab delta 
brought about also a migration of the population to the south. The latest Bronze 
Age sites are the settlements of Takhirbay 3,1 dating from the second half of the 
2nd millennium BC (Masson 1959). 

Somewhat later than in Margiana and southern Bactria there forms a culture 
in northern Bactria (Map 17). In southern Uzbekistan it is represented by the 
sites of Sapalli, Dzharkutan, Mollali, Bustan and others. A. Askarov 
distinguishes them as the Sapalli culture. In actual fact, by all the culture- 
determining characteristics they belong to the Oxus culture and may be viewed 
only as its local variant. Based on the stratigraphy and ceramics the stages of 
Sapalli and Dzharkutan have been assigned to 1700-1300 BC and those of 



THE AGRICULTURAL TRIBES OF SOUTH CENTRAL ASIA 227 

Kusali, Mollali and Bustan (1300-900 BC) to the Final Bronze Age (Askarov 
1977: 101; Askarov and Abdullaev 1983: 40-44; Askarov and Shirinov 1993). 
This chronology appears to require refinement and shifting to an earlier time. 

The Oxus culture is the pinnacle of the development of the region's Bronze 
Age farming civilization. Here were uncovered settlements with multi-room 
houses and — constituting the peculiarity of the culture — square, rectangular and 
round fortresses with towers and by-pass corridors (Kelleli, Sapalli, Dashly 1, 3), 
temple complexes with altars related to the fire cult and hallucinogenic drinks 
(Togolok 21,3, Dzharkutan), and palaces with ceremonial halls (Gonur, Dashly 
3). 

The staple of the economy was irrigation farming supplemented by cattle- 
breeding whose role grew towards the late 2nd millennium BC. Great strides 
were made by craft specialists: there were wineries, workshops for 
manufacturing unornamented, light colored, wheel-made ceramics of varied 
types; workshops for manufacturing figurines, stone articles, including 
cylindrical and flat seals. Metal-working became particularly developed. Ore 
came from Iran and Afghanistan (Rusanov 1982). Production included 
implements, weapons, vessels bearing depictions of hunting scenes and animals, 
polymorphous creatures, as well as seals with cross-shaped patterns and images 
of goddesses, animals, snakes and dragons (Sarianidi 1998). 

Considerable changes occurred in ideology. Four types of burials were 
established: 1) under the house floors; 2) in the deserted areas of a settlement — 
these rites maintain the ancient tradition of Anau; 3) in a burial ground in a pit, 
often faced with bricks — they are predominant in Margiana, whereas in Bactria 
catacomb burials prevail (4). The dead lie flexed on their side, accompanied with 
vessels, sometimes with a ram's carcass and ornaments. Very rich burials stand 
out; some of these are cenotaphs, sometimes containing a ram or a clay figure. 

The architecture of the temples and applied art reflects the religious system 
with complex rites and the cult of a female deity. The BMAC mythology forms 
part of the circle of ancient Near Eastern beliefs and, judging by the artistic 
images, is closest to the pantheon of Elam (Amiet 1977, 1997; Francfort 2001; 
Litvinsky 1989; Klochkhov 1997; Antonova 2000). 

Contrary to the opinion of many western scholars, the BMAC continued to 
develop in the second half of the 2nd millennium BC. It is exactly from this time 
that we find many sites in Uzbekistan, assigned by Askarov to the Dzharkutan, 
Kuzali, Mollali and Bustan stages. In this period active contacts with the 
Andronovans were established. 

At the late stages of the Oxus culture this population migrated eastwards and 
formed the BMAC in the foothill valleys of Tadzhikistan. The specific 
ecological conditions predetermined the peculiarity of the cultural development. 
In the narrow mountain valleys only the bogar type of farming and cattle- 
breeding were possible, while the rich pastures of the south would enable cattle 
to graze throughout the year (P'yankova 1998: 163). Therefore, in Tadzhikistan 
there formed two types of cultures: cattle-breeding ones and the Tadzhik variant 
of the BMAC. 

The culture is represented by the settlements of Kangurt-Tut, Teguzak, 
Dakhana in the valley of the Vakhsh river and the summer site of Baraki-Kuduk, 
as well as the cemeteries of Kangurt-Tut and Nurek and in the Hissar valley by 



228 CHAPTER SEVENTEEN 

the Tandyryul and Zarkamar cemeteries, a grave at Tup-Khona, by ceramic finds 
in Kafirnighnan and Dushanbe (P'yankova 1994: 355-372; 1998: 163-170; 
Vinogradova 1994: 29-47; 2000: 89-109; 2001: 199-201; Gotzelt et al. 1998: 
115-144) and also by the layers of stage B at the settlement of Shortughai in 
Afghanistan (Francfort 1989). 

According to the stratigraphy of Shortughai, the BMAC period containing the 
ceramics of the Fedorovo type is chronologically subsequent to stage A of the 
Harappan culture and is cut through by the burials of the Bishkent-Vakhsh 
culture, synchronous with the Andronovo culture ceramics of the Fedorovo type 
and the late Andronovo ceramics with applied-rollers (Table 13). At the 
settlement of Teguzak the BMAC layer lies over that of the Neolithic culture of 
Hissar (Vinogradova 2000: 93-94), while at Kangurt-Tut the sequence runs: 
Hissar-BMAC-Yaz I, In France (CNRS) and Russia (GIN) radiocarbon dates 
have been obtained: 1594 BC and 1291 be and 1756 BC and 1320 be 
respectively. 

The settlements are located on the mountain slopes. The houses have stone 
foundations for the walls, occasionally the floor is faced with pebbles, and round 
hearths; they are organized in terraces on the slopes. There are also proto-yurts 
faced with stone. The economy was mixed. The bogar type of farming is 
documented by barley and wheat, querns, and grain-storage pits. They reared 
cattle (52%), ovicaprids (19.5%), horses (19%), donkeys (5.6%) and camels. The 
stone articles continue the tradition of the Neolithic Hissar culture. The 
development of metalworking is evidenced by finds of molds. A substantial 
percentage is accounted for by bronze implements of the Andronovo type. 
Ceramic production is highly developed. The vessels are wheel-made, fired in 
two-tiered kilns and belong to types of the Mollali and Bustan stages of the 
Sapalli variant of the BMAC. The Mollali stage at the settlement of Dakhana 
yields 86% wheel-made pottery, and 14% hand-made. Andronovo-type vessels 
are absent. In Kangurt-Tut wheel-made pottery of the Mollali and Bustan stages 
accounts for 53%, hand-made for 47%; the number of Andronovo pieces is 
insignificant. But in Teguzak Andronovo vessels total as much as 43%. 

The burial rite is typical of the Mollali stage. Dismembered and children's 
burials are known within the settlement. In the Kangurt-Tut and Tandyryul 
cemeteries catacombs were discovered sealed with stone or clay. Of frequent 
occurrence are cenotaphs, some of them containing a clay anthropomorphic 
figure. Also earth pits and pits with a step are found. The dead were in the flexed 
position on their side. There are also well burnt bones and ash. At Teguzak a 
ritual hearth, coated with clay, was found, which testifies to the cult of fire. The 
graves contained wheel-made and, rarely, hand-made vessels, beads of lapis 
lazuli, ornaments and miniature votives (knife, mirror, ax-adze, razor, pitchfork). 
These have analogies in Sapalli. Thus the Tadzhik variant, though belonging to 
the BMAC, is marked by a peculiar economic and cultural type. Of great 
importance were the contacts of Tadzhikistan's farmers with the neighboring 
Andronovans (see Chapter 21). 



CHAPTER EIGHTEEN 
THE OCCUPATION OF CENTRAL ASIA BY PASTORAL TRIBES 

The first Indo-Aryan migration to the south 

Those who support an Indo-Iranian or Aryan attribution to the Andronovo 
culture have provided evidence of the migration of steppe tribes to the south into 
Central Asia and Afghanistan in the second half of the 2nd millennium BC (see 
bibliography in Kuz'mina 1994; P'yankova 1998). The appearance of horse 
bones and the depiction of horses in Turkmenia, where the horse was earlier 
unknown, and the spread of horse-drawn chariots in Namazga VI are all 
connected with this wave of migration (Kuz'mina 1980). In Bactria one horse 
burial is known from Dashly 19 (Sarianidi 1977: 148) and there are images of 
horse heads on ceremonial bronze axes and handles (Fig. 95; Pittman 1984: fig. 
32; Amiet 1988: fig. 9b; Ligabue et al. 1988: figs. 96, 101; Fig. 95). 

Recently there has emerged evidence of contacts between the population of 
the Urals and the south already in the Novokumak horizon. A lapis-lazuli bead, 
originating from Bactria, has been found in a Sintashta settlement; a plate imita- 
ting pottery from the BMAC has been found in a Petrovka layer on the settle- 
ment of Ust'e (Vinogradov 1995b: 72); under the mound of a kurgan at Krasnoe 
Znamya (Fig. 60) a bronze mirror with a protruding handle has been uncovered, 
ceramics of the Sintashta-Abashevo type, and two horse skulls; in the main 
burial (nr. 1) on this site was found a spear, adze, chisel, awl, knife, abrasives, 
three Sintashta vessels, bones of a ram and a dog skeleton (Sungatov et al. 1995: 
60, fig. 2). The mirror is of Bactrian type (Sarianidi 1977: tab. 2.8, fig. 40). 

The Zardcha-Halifa grave at Panjikent near Sarazm, on the left bank of the 
Zeravshan (Bobomulloev 1993; 1997; Bostongukhar 1998) is of great impor- 
tance for clarifying the time and fate of the early Andronovo tribes (Fig. 65). The 
deceased lies in an oval grave, 3.1m long and 3.5m deep. He lies flexed on his 
right side, head to the south-west, one arm under the head and the other on the 
stomach. The skeleton of a ram was placed at his head. This rite is typical of the 
Sapalli culture of northern Bactria (Askarov 1977: 138). The assemblage of 
grave goods was rich. Ceramics included globular vessels with a narrow neck 
two examples of an incised base and a tamga on the shoulder of one of the 
vessels. These were wheel-made and fired pink. Such ceramics are comparable 
to the Dzharkutan stage of the Sapalli culture, e.g., the Dzharkutan cemetery, 
partially Dashly 3 (Askarov 1977, figs. 31, 32; Askarov et al. 1983: 7, tab. xxi.7, 
xxviii; Sarianidi 1977, figs. 27, 28). 

There are also analogies with other assemblages in the BMAC and in Iran. In 
Sapalli, Hissar III and the BMAC there are parallels to the bronze vessel 
(Askarov 1977, tab. xxvii.15; Sarianidi 1977, fig. 41.9; Amiet 1988, fig. 3), a 
bronze vessel with a knurled neck found in Bactria (Amiet 1988, fig. lib, d); a 



230 CHAPTER EIGHTEEN 

temple ring with bulges at the ends is similar to ornaments from Sapalli (Askarov 
1977, tab. xxxix. 17, 19) along with a gold cup, razor -knife, hafted dagger with 
straight shoulders, and gold and turquoise beads. A stone phallus-shaped pestle, 
26.5cm long and 4.5cm in diameter is assigned to the same complex. N. 
Boroffka (1998: a, b, 25) attributed it to the Andronovo V B type, close to type 
IV, which is characteristic of Central Asia during the Namazga V-VI period, and 
he dated it to 1800-1400 BC. Comparison with examples from Ulug-Depe and 
Parhai makes it possible to propose a date of 1800-1600 BC. 

A bronze pin is of special interest (Fig. 68: 1). It is 18cm long, crowned with 
the figure of a horse. Pins with zoomorphic heads are widely known at Sapalli 
Tepe, Dzharkutan, Dashly-3, burials in Bactria, Hissar III and the Khak hoard 
(Kuz'mina 1966, tab. xvi; Askarov 1977, tab. xl.l, 1.6, 5; Sarianidi 1977, fig. 43, 
44; Sarianidi 1988, tab li, lvi; Askarov et al. 1983, tab. xxi.l; Amiet 1988, fig. 
12; Ligabue et al. 1988, fig. 83). However, I am not aware of a horse-headed pin 
among the whole range of farming cultures. Stylistically, this image resembles to 
some extent the images of a horse found on a gold temple ring from the 
Andronovo cemetery of Mynchunkur (Kuz'mina 1994: 256, fig. on p. 5), on a 
knife from the Seyma cemetery (Bader 1970, fig. 52). The static posture, 
exterior, long tail and depiction of the mane are all similar. 

Fragments of shield-shaped cheek-pieces were found at Zardcha-Halifa (Fig. 
66), some of which have been reconstructed. These are of bone, 8cm in diameter, 
with a large central slot surrounded by a ridge and four solid tenons 
(Bobomulloev 1997: 127a, b, 4: 1, 2). They are assigned to type 1 of my 
classification and represent the most archaic form; they are characteristic only of 
the early Sintashta complexes of the Urals: Sintashta (Gening et al. 1992, fig. 
57.8; Kuz'mina 1994: 171-189, tab. 4, fig. 37) and Bol'shekaragan (Botalov et 
al. 1996: 80, 81, fig. 17.10, 18.4). In Tanabergen (Fig. 64; Tkachev 1998, figs. 2, 
10, 11) there are cheek-pieces of type 1 but without the ridge; in the Potapovka 
burial on the Volga they have the ridge but with additional slots (Vasil'ev et al. 
1994, fig. 33.1, 42.3). A pair of bronze bits with rings and couplings at both 
ends, 11.5 and 12cm long, is connected with the chariot complex. Attempts to 
compare them with a bit from Kairak-Kum (Bobomuloev 1997) do not seem to 
be correct and I am unaware of any analogies. 

Analysis of the Zardcha-Halifa material indicates that the greater part of the 
art may be assigned to the BMAC while the horse -headed pin and the cheek- 
pieces are characteristic only of the early Sintashta sites of the Urals. They help 
establish the zone where horses and chariot first appeared in Central Asia and the 
Zardcha-Halifa burial emerges as a reflection of the first wave of Indo-Iranian 
migrations to the south. 

Of great importance is the discovery of the metal-working settlement of 
Tugai near Samarkand near the polymetallic deposits of the Zeravshan ridge 
(Avanesova 1996). Archaeologists uncovered a semi-subterranian house and 
metal-working complex with round hearths, furnaces for smelting ore, and traces 
of metal -working such as ore, bars, coal, and clay crucibles, as well as a bronze 
celt, stone axe, hammer and arrows. In addition to metal-working there is 
evidence for stock-raising as bones of cattle, sheep and goat were recovered. The 
cultural attribution of the complex is established by its ceramics which comprises 
22 hand-made vessels (Fig. 69). They are of the Petrovka type, gray and black, 



THE OCCUPATION OF CENTRAL ASIA BY PASTORAL TRIBES 23 1 

with ornament made by toothed stamp, sometimes in a caterpillar pattern or with 
a comb. Ornamental motifs include zigzags, triangles and often herring-bone at 
the base. Two vessels have traces of textile impressions. The fabric includes 
admixture of shell; two vessels have talc as an opening agent (Avanesova 1996: 
122, fig. 43, 44). 

The discovery of the Tugai settlement verifies the early wave of Andronovo 
migrations. The presence of talc, specific for the Urals, in the ceramics indicates 
the origin of the population who were searching for new sources of ore. 

The date of the site can be established on the basis of six quality vessels 
found in a closed complex (the house). These comprise conical or semi- 
hemispherical cups, one of them black polished; others are red polished. These 
ceramics were brought from the neighboring settlement of Sarazm which is 
situated 27km from Tugai. N. A. Avanesova (1996: 120, fig. 41) assigned the 
material to the fourth layer. The site of Tugai makes it possible for the first time 
to synchronize the pastoral early Andronovo culture with the settled farming 
culture of Sarazm which had wide contacts with western Asia, especially 
Baluchistan and the Indus valley. 

What is the absolute date of these sites? Chronological issues that form the 
basis of any historical constructions are currently at the center of heated dis- 
cussion (Kuz'mina 1998). In Russian science the Bronze Age chronology has 
been established using the scale of Mycenae that has been developed in western 
scholarship; the steppe sites have been cross-dated to Mycenae according to the 
analogical method. In western archaeology, however, especially in that of the 
Near East, calibrated radiocarbon dates have provided the basis since the public- 
ation of Colin Renfrew's "Wessex without Mycenae" in 1968. 

The basis of the chronology of the sites of the Novokumak horizon has been 
built on a series of synchronizations: 1) according to cheek-pieces and ornament 
from shaft grave IV at Mycenae, dating 1570-1550 BC, which provides a 
terminus post quern for the cheek-pieces of type 1; this establishes a date for the 
horizon of the 17th 16th centuries BC (Smirnov and Kuz'mina 1977: 40-50); 2) 
according to plaques from the Monteoru culture (Litvinenko 1996; Matveev 
1996), and 3) according to segmented faience beads that are found in a number 
of European cultures. 

Mycenae has been dated within a century on the basis of its synchronization 
with Egypt and the Near East, the revival of the schemes of G. Karo and A. 
Furumark and the date of the volcanic eruption of Santorini established by 
radiocarbon dates. There is also a tendency to adjust the dates of the traditional 
chronology of the sites of Central Europe, the Reinicke Al and A2 stages, on the 
basis of dendrochronology (Krause et al. 1989; Kroemer et al. 1993; Randsborg 
1992; Kuniholm 1993; 1996). The dendrochronological adjustments worked out 
for Europe advances the traditional dates by one or two centuries but they still 
remain younger than the calibrated radiocarbon dates. Calibrated dates differ 
significantly from the historical chronology of Egypt and the Near East 
(Chernykh 1997) and there is considerable variation in their usage (e.g., the 
Sintashta cemetery dates 2250-1390 BC without calibration). 

A series of new radiocarbon dates for the cemeteries of Krivoe-Ozero, 
Potapovka and Utevka VI places the Novokumak horizon at the turn of the 3 rd- 
2nd millennia BC (Vinogradov 1995; Anthony and Vinogradov 1995; Kuznetsov 



232 CHAPTER EIGHTEEN 

1996; Trifonov 1996). Dates have also been received from Siberia (Kiryushin 
1991; Orlova 1995; Matveev 1998); however, not all accept these new dates. 
There are ten radiocarbon dates for the settlement of Arkaim (two from the 
University of Arizona) and nine more from other sites of the Urals. According to 
G. Zdanovich (1997: 60), "the main zone of reliable dates corresponds to the 
18th— 16th centuries BC although there is another group, for Krivoe Ozero, that 
dated to the 21st-20th century BC." The date of fragments of a chariot from a 
grave at Satan is 1557-1255 BC (Novozhenov 1994: 160). The divergencies in 
dates do not permit a single conclusion. 

The differences between the chronological systems for Central Asia proposed 
by Russian and foreign archaeologists are especially marked. The chronology of 
the Turkmenian sites has been worked out by V. M. Masson and his scheme is 
universally accepted. V. I. Sarianidi (1977: 158) assigns the completion of the 
BMAC in Afghanistan to the middle of the 2nd millennium BC. For Margiana he 
(1990: 5) proposes the following scheme: 1) Kelleli- 19th/ 18th centuries BC; 2) 
Gonur-17th/15th centuries BC; Togolok-15th/9th centuries BC. A Askarov 
(1997, fig. 31) dates the northern Bactria periods as follows: 1) Sapalli stage- 
1700/1500 BC; Dzharkutan-1500/1350 BC; Mollali-1350/1000 BC (earlier he 
dated the Dzharkutan period to the 3rd quarter of the 2nd millennium BC, Kuzali 
to the 13th/12th centuries BC, and the Mollali period to the llth/lOth centuries 
BC (Askarov et al. 1983: 33, 39, 42, 44). L. B. Kircho and S. G. Popov have 
insisted on the traditional dates (1998) but V. M. Masson (1999) accepts the new 
calibrated chronology. 

This chronology has been severely criticized by western scholars who insist 
on dating the sites approximately 300 to 500 years earlier and who date the 
BMAC to 2300-1500 BC (Francfort 1989: 241, 242; Hiebert 1993: 1994: 75-87; 
Lyonnet 1996: 16, 67; Gotzelt 1996). 

Regarding the settlement at Tugai, B. Lyonnet (1996: 60, 68, 120) disagrees 
with N. A. Avanesova and accepts the early date of the ceramics of the Sarazm 
type to the 3rd millennium BC. A. I. Isakov (1991: 113, 115) assigns the fourth 
Sarazm layer to 2300-1900 cal BC. The burial at Zardcha-Halifa has been 
assigned by S. Bobomuloev (1993: 63) to the early stage of Namazga VI and to 
the Dzharkutan stage of the Sapalli culture which A. Askarov dates to 1700-1500 
BC (radiocarbon dates of 1650±60 BC). Later he (1997: 132) compared the 
complex with Shah-Tepe Ha and Hissar III and accepted a calibrated date for the 
Dzarkhutan temple of 2034-1684 cal BC; he assigned Zardcha-Halifa to 2100- 
1700 BC. 

Thus, the date of the sites of the Novokumak horizon coincides on both the 
European and Central Asian scale: the traditional chronology sets it to the 17th- 
1 6th centuries BC according to synchronization with Mycenae while the redated 
Mycenae and the dendrochronological evidence places Bronze Age A2 to the 
18th— 17th centuries BC. The calibrated radiocarbon dates places their age in 
both Europe and Central Asia to the 2 1st- 18th centuries BC. This last date is also 
supported by a new radiocarbon date for a cheek-piece from Monteoru (Zaharia 
1990: 43). 

The acceptance of the calibrated radiocarbon dates resolves the issue of the 
chronological priority of chariots between the Near East and the steppe as they 



THE OCCUPATION OF CENTRAL ASIA BY PASTORAL TRIBES 233 

indicate the importance of the Sintashta sites in spreading the horse-drawn 
chariot in the Old World. 

The discovery of the settlement at Tugai and the burial at Zardcha-Halifa 
allows us to synchronize for the first time the settled farming sites of the Sarazm 
IV type and the BMAC of the Dzharkutan period with early Andronovo. It 
supports the model of a migration of a group of steppe tribes from the Urals to 
the south where they carried their metallurgical skills, horses and cheek-pieces 
for harnessing chariots. This is all in accord with the hypothesis that the 
Andronovo culture was Indo-Iranian and it reflects an early stage in the 
migration of Aryan tribes to the south. 

What is the further fate of these migrants? I. M. D'yakonov (1960) postulated 
the idea that part of the Indo-Iranians left their European homeland and settled in 
central Asia. The same idea was independently developed by Th. Burrow (1973) 
who proceeded from the observation that there are many loans from Dravidian 
and aboriginal languages in India in the Vedic language which are absent in 
Iranian. It follows that the Iranians replaced a related Indo-Iranian population 
who had come earlier. This hypothesis has been energetically supported by A. 
Parpola (1988). During the Arkaim discussion in 1999, C. Lamberg-Karlovsky 
stressed that traces of the Andronovo migration and their interaction with 
farming populations are found in Central Asia but do not extend to India while 
there is clear evidence of the active influence of the BMAC in Baluchistan and 
Pakistan which he labels the Oxus culture (Lamberg-Karlovsky 1994; 1996). 
This idea must be considered in the light of data from the Samarkand burial 
where the assimilation of steppe cultural elements with local farmers is evident. 

However, Vedic texts do not support the notion that the Indo-Aryans who 
arrived in India were representatives of the Oxus culture. They were pastoralists, 
serving as nomads for half the year, moving in their vehicles with their herds or 
walking according to the asvamedha rite after the king's sacred horse to conquer 
new lands (Rau 1983; Elizarenkova and Toporov 1995). In his article "Is Vedic 
archaeology possible?", W. Rau (1977) wrote that archaeologists could count on 
finding only temporary camps with a thin cultural layer and a handful of sherds 
of hand-made pottery in India. Even among the Persians who came to Iran in the 
1st millennium BC, only six tribes settled down while four tribes remained 
nomadic (D'yakonov 1956: 187; Grantovsky 1998: 111-113). The economic and 
cultural types of the Aryans corresponds only to the material culture of the 
steppe pastoralists, especially those of the Andronovo culture. 

The first wave of migrating Andronovans reached the Zeravshan already 
during the Novokumak stage. It is not known at present whether they continued 
further and entered India, coming into contact with the Harappan culture. It is 
also not clear whether they spoke a common Indo-Iranian language or whether 
they were divided into Nuristani, Indo-Aryans and proto-Iranians. In any case, 
judging from the similarity of their religious and mythological ideas, the cult of 
the horse and the chariot, the strata of elite chariot -warriors and the common set 
of weaponry, the culture of the individual Indo-Iranian tribes was still very 
similar by the 17th century BC when those Aryans neighboring the Hurrians 
appeared in Mitanni. 



234 CHAPTER EIGHTEEN 

The second stage of Andronovo migration 

The second stage of Andronovo migration comprises the 15th 13th centuries 
BC. The synthesis of different cultural components led to the unification of 
culture during the Petrovka stage. The large depositions of sacrificed horses and 
other animals and wealth disappear, burial rites become simpler and there is a 
standardization of metal and ceramic technology and ornament. Tribes settle in 
northern and central Kazakhstan while some groups reach Central Asia. 

The Alakul' type was formed on the basis of the Petrovka type. The evidence 
of ceramics suggests that the Fedorovka-type sites emerged simultaneously in 
central and eastern Kazakhstan. Further development appears to involve a 
complex and difficult process of interaction between these two groups of people 
with one another and with neighboring alien cultures. Assimilation and 
integration led to the cultural unity of the Andronovo culture. The unity of the 
Timber-grave culture was formed in the forest-steppe on the basis of the 
Poltavka culture with influence from the Catacomb and Abashevo populations 
(Merpert 1985; Merpert et al. 1985: 10-28), and in the steppe the Catacomb 
culture was a major component (Malov et al. 1995; Otroshchenko 1990). 

Once the situation had become stable and the Timber-grave culture was 
unified in the European steppe while the Andronovo culture was unified across 
the Asian steppe, there was a push for new territories. The Timber-grave people 
moved west along the river valleys deep into the forest zone, occupying a 
considerable portion of the Ukraine (Berezanskaya 1972; 1978; Berezanskaya et 
al. 1986) and the Crimea. A separate wave of Timber-grave people also moved 
to the south-east. 

The Andronovans managed to master enormous territories (Map 9). They 
pushed north to the forest zone (Sal'nikov 1967), and settled right across the 
whole of Kazakhstan (Sorokin 1966; Margulan et al. 1966). The Alakul' tribes 
migrated from the west to central Kazakhstan which pushed a part of the 
Fedorovo population eastwards into eastern Kazakhstan and further to the Ob 
and the Minusinsk Basin. This route of the Fedorovo people is indicated by 
ceramic analysis: the purest complexes are found in central Kazakhstan. There is 
a group of cemeteries, similar in both rite and ceramics to those of central 
Kazakhstan but now in eastern Kazakhstan. Sites of the Ob region resemble 
those of eastern Kazakhstan; on the periphery along the taiga zone and the 
Yenisey there is the strong influence of local aboriginal cultures. Within the 
main Andronovo territory there was an intensive process of integration between 
the Alakul' and Fedorovo populations and this is seen in sites of mixed type 
(Kuz'mina 1985a). In the Urals a wide zone of mixed Timber-grave and 
Andronovo sites is formed (Chlenova 1981). The Andronovo influence spread as 
far as the Don and Dnieper region (Berezanskaya and Gershkovich 1983; 
Otroshchenko and Motsya 1989; Posrednikov and Kravets 1992). 

Post-Soviet Central Asia, which is the center of the Eurasian continent, is 
extremely diverse in its natural and climatic conditions. It embraces four zones: 
1) the steppes that extend northwards; 2) the Kyzyl -Kum and Kara-Kum deserts 
that occupy the central part; 3) the high Tian-Shan and Pamir mountains which 
are situated on the east and south-east; 4) in the south-west the low Kopet-Dag 
mountains that separate Central Asia from Iran, and along their northern slopes 



THE OCCUPATION OF CENTRAL ASIA BY PASTORAL TRIBES 235 

there is a narrow strip of fertile lands; oases are also located at the mouths of the 
Tedjen and Murgab rivers and downstream on the Zeravshan. 

The diversity of the ecological conditions, which changed many times over 
the course of the Eneolithic and Bronze Age, predetermined different directions 
of economic and cultural development and different orientations of the ethno- 
cultural relations of the regions involved. The northern and eastern areas of 
Central Asia formed part of Eurasia's steppe cultures, The southern areas 
belonged to the range of the cultures of the Ancient Near East. Central Asia, 
serving as a bridge between these two worlds, was the zone where contacts were 
realized between tribes belonging to different economic and cultural types and 
where, over the course of the whole historic epoch, lay the pivotal route of ethnic 
migrations, the road leading to Iran and India. 



CHAPTER NINETEEN 

THE SETTLEMENT OF PASTORAL TRIBES IN CENTRAL ASIA 

While the south saw the development of the farming cultures of Anau and the 
BMAC, in the north in the steppes the Andronovo culture and its representatives 
established contacts with the farmers and gradually began their advancement 
southwards. We can discern three stages of the migration of the pastoral steppe 
tribes to Central Asia: stage I - Sintashta-Petrovka; stage II - developed Timber- 
grave and Andronovo of the Alakul', Fedorovo and mixed types; stage III - late 
Timber-grave and Andronovo with applied-roller ceramics. The number of the 
northern sites invariably grows and in the Final Bronze Age (13th)12th-9th cen- 
turies BC they occupy the whole territory of the region. 

The Timber-grave and Andronovo tribes of the Alakul' type emerged in the 
late second and third quarters of the 2nd millennium BC in the Urals and western 
Kazakhstan, while the Fedorovo type tribes emerged in central and eastern 
Kazakhstan. They were extremely expansive which was conditioned by the 
peculiarities of the steppes' ecology and their economic and cultural type. The 
exhaustion of the pastures around a settlement would compel its inhabitants to 
change their location every 25 years. The pressure of the excess population on 
the limited resources of the steppes would require them to develop new 
territories. The Fedorovo tribes advanced into Siberia. In Kazakhstan the 
processes of active integration among the tribes and the formation of mixed 
types of sites were underway. But the principle direction of movement was into 
those regions of Central Asia that had already experienced cultural interaction 
and assimilation of the steppe tribes with both each other and with the southern 
farmers. This led to the formation of numerous and quite original types of sites. 
Their mixed character presents difficulties in their classification and causes 
debate among scholars as to their attribution. 

The sites of the steppe population are united under the terms Andronovo 
(Chernikov 1957: 30; Zadneprovsky 1966: 213; Gryaznov 1970: 40; Askarov 
1962a: 3,17; Gulyamov et al. 1966: 187, 213; Avanesova 1979; 1991); Taza- 
bagyab-Andronovo (Masson 1959: 116,117; Avanesova 1985); of the steppe 
type (Kuz'mina 1964a: 147, 154; 1988: 35, 36; Mandel'shtam 1966: 242-243; 
Masson 1966: 208, 261; Itina 1977a: 232). It is advisable to apply the last term to 
assemblages from wind-eroded sites where the ceramic complex is not infor- 
mative enough and does not contain diagnostic material. It should be emphasized 
that all the ceramics of Central Asia with the exception of the Fedorovo type are 
poorly ornamented, the decoration being usually located in one zone (over the 
shoulder), more rarely in two zones (over the neck and shoulder). This choice of 
zones reflects the traditions of either the Petrovo-type sites or the Timber-grave 
culture and differs from the Alakul' principle of placing ornament on the rim and 
shoulder with a gap on the neck and the Fedorovo one of decorating the rim, the 



238 CHAPTER NINETEEN 

shoulder and the neck. Late Fedorovo pottery is marked by poorer ornament and 
a shifting of the ornamented zones. In accordance with the statistically stable 
aggregate of the dwelling types, the burial rite and the ceramics one distinguishes 
in Central Asia several independent cultures and types of monuments of the 
Andronovo cultural community. 



Tazabagyab culture: The Aral Sea Littoral variant 

This culture was discovered to the south of the Aral Sea in the desert over the 
ancient dried-up river-beds of the Amu-Darya (future Khorezm). Around fifty 
settlements have been discovered: Kavat 3, Angka 5, Bayram-Kazgan, Kokcha 
15, 15a, 16, Dzhanbas 21 and the Kokcha 3 cemetery (Tolstov 1948, 1962; Itina 
1961, 1967, 1977, 1978; Vinogradov et al. 1996). Large settlements are absent; 
the dwellings are located in twos and threes among the fields. The house is semi- 
subterranean of timber-frame type measuring 7-12 x 10-14 m. In the center there 
is a square (more rarely circular) hearth with clay sides, storage pits and querns. 
The entrance is a corridor-ramp leading to the open terrace. 

As distinct from the other steppe cultures, Khorezm' s economy was based on 
irrigation farming. The 150-200m long canals would irrigate small rectangular 
fields (Andrianov 1969). Cattle and particularly ovicaprids, horses and Bactrian 
camels were reared. The short-term sites in the sands point to a mobile type of 
cattle-breeding. The use of wheeled transport is evidenced by clay models of 
wheels. The source of ore was the Bukan-tau and Tamdy-tau mountains, where 
ancient workings and copper-smelteries were discovered (Itina 1977: 136, 137). 
The findings of a special ladle for pouring metal and stone molds are indicative 
of the household exchange type of metalworking. With regard to the types of 
articles, particularly adzes and figure-of-eight temple rings, Khorezm belonged 
to the western Andronovo metallurgical province and was especially close to the 
Elenovka-Ushkatta center. 

The ceramics of Khorezm are hand-made and represented by the pots of the 
Andronovo type with a rounded shoulder or a ledge, the Timber-grave biconical 
pots, jars and specific vessels with a globular body and a narrow neck. Sixty 
percent of the pots are decorated with an ornament executed in indented (20%) 
or plain stamp and incision. Apart from the Andronovo and Timber-grave 
elements of the geometrical ornament, there are specific open triangles, and 
triangles with a fringe. 

The cemetery of Kokcha 3 contains around a hundred burials in earthen pits 
(there were no kurgans). The dead lie flexed, their head to the west, men on their 
right, women on their left side. There are double burials of mixed sex. At the 
head stands one, rarely, two vessels; women occasionally wear bracelets, temple 
rings, beads. The cults of Khorezm are also evidenced by figurines of the horse 
and camel. 

The cultural attribution and origin of Khorezm's sites are disputed. A. Aska- 
rov (1962a: 3,17; 1966: 187, 213); Gulyamov et al. (1966); Zadneprovsky (1966: 
213), M. Gryaznov (1970: 40), and N. Avanesova (1962: 57-59) assign these 
sites to the Andronovo culture. On the other hand, S. Tolstov (1962: 57-59) and 
M. Itina (1977: 139, 140, 176) regard them as a special culture formed from the 



THE SETTLEMENT OF PASTORAL TRIBES IN CENTRAL ASIA 239 

migration of the Timber-grave and Andronovo tribes, whose synthesis had taken 
place already in the Urals. On arrival in the fertile lands of Khorezm the steppe 
tribes entered into interaction with the farming population — representative of the 
indigenous culture of Suyargan. The existence of a distinct Suyargan culture now 
appears insufficiently substantiated, yet the general scheme is correct. The 
legitimacy of distinguishing the sites of Khorezm as a particular Tazabagyab 
culture is borne out by the statistically stable combination of characteristics 
marking it off from the Timber-grave and Andronovo cultures. In architecture 
we find the absence of large settlements and a special type of dwelling; in the 
burial rite it is the absence of burial mounds, stone and wooden structures, 
placement of men on their right side; in the ceramic complex it is a mixture of 
the Timber-grave and Andronovo forms of vessels, the specific type of pots with 
a globular body and a narrow neck, local types of ornaments, and, finally, it 
offers an absolutely peculiar economic and cultural type with irrigation farming, 
which is well -suited to the ecological conditions of Khorezm. 

The Tazabagyab culture has been dated to the 15th— 1 1th centuries BC. The 
upper date is determined by the time of the spread of the Amirabad culture, 
marked by applied-roller ceramics. In light of new data the genesis and 
chronology of the Tazagabyab may be refined. Typical of Khorezm, the 
biconical vessels, the bizonal application of the ornament, the prevalent patterns 
of herring-bone and opposed triangles, ornamentation by cones and oblique 
triangles, and ornament on the base are most fully analogous to the ceramic 
complex of the Petrovka-type sites in the Urals and western Kazakhstan, whence, 
presumably, came the migration. This enables us to revise the formative period 
of the culture to an earlier date. 

The complex genesis of the formation of Khorezm' s population is corrobo- 
rated by the anthropological data. Most of the skulls belong to the population 
type known from the Trans-Volga and western Kazakhstan, where there are 
skulls close to the types of southern Turkmenia and Harappa. The series is 
extremely mixed which indicates the processes of assimilation (Trofimova 1961; 
Ginzburg et al. 1972: 86-88). 

Located in alternating strips with the Tazabagyab sites, there are the summer 
sites of the Andronovo pastoralists, Dzhanbas 34 and Kokcha 19, with round 
dwellings (proto-yurts) and the ceramics of the Kozhumberdy and Sol'-Iletsk 
types. The same pottery overlies the Tazabagyab irrigation canal and cultural 
layer at the settlements of Kokcha 15, 16 (Itina 1977: 52, 57, 58, 79-82; 104-109; 
1 19-121 ; fig. 22-24, 39, 40, 57, 59, 61). We thus may synchronize the developed 
stage of the Tazabagyab culture with the 2nd stage of the Andronovo culture of 
the third quarter of the 2nd millennium BC. 

The population of Khorezm continued to maintain contacts with the farmers 
of the BMAC. At the settlements of Kokcha 15 and 15a and others there were 
found fragments of the light-clay wheel -made vessels with a conical bottom 
typical of the Namazga VI stage (Itina 1977: 69, 72, 193, fig. 18: 8; 1978: 525). 
In addition, of southern origin are the types of pin with a double-spiral head, 
earrings with cones, and clay figures. These findings afford a unique opportunity 
to synchronize the Tazabagyab culture with the culture of the ancient farmers, 
and, through it, also the Andronovo, and establish the presence of wide cultural 
ties of the population of Khorezm. 



240 CHAPTER NINETEEN 

The Tazabagyab culture: The Lower Zeravshan variant 

In the Kyzyl-kum desert north of Bukhara alongside the delta of the Zeravshan' s 
Makhan-Darya and Lakes Gudzhayli-Gurdush and Tuskan, now desiccated, and 
in the sands of Kaptarnikum thirty sites and the cemeteries of Kyzylkyr and Gur- 
dush (Gudzhayli) have been discovered (Figs. 45: 2-4, 6, 8, 11; 46). At Kashka- 
Dar'ya four sites are listed, at Dar'yasai — the Madami site (Gulyamov 1956: 
149, 156; Gulyamov et al. 1966; Askarov 1962a, b, 1964, 1965; Kuz'mina 1968; 
Duke 1969; Avanesova 1985). All the sites are wind-eroded. The cultural layer is 
preserved at the sites of Paykent 6, Gudzhayli 9, and Bol'shoy Tuskan 3. The 
type of dwellings is not identified. Hearths/fires have been discovered, in Mada- 
mi — stone hearths and a copper-furnace. Yu. Gulyamov (1956) noted the simil- 
arity with those of the Tazabagyab culture. A. Askarov (1962a: 3, 17; Gulyamov 
et al. 1966: 187, 213, map), Yu. Zadneprovsky (1966: 213), and N. Avanesova 
(1979, 1985, 1991) assign them to the Andronovo culture. M. Itina (1967: 76, 79, 
1977: 232) and E. E. Kuz'mina (1988c: 307) substantiated their attribution to the 
Tazabagyab culture. 

Just as in Khorezm, the economy is based on irrigation farming (canals have 
been discovered, querns are present). Ovicaprid bones have been found. The 
types of metalwork belong to the western Andronovo metallurgical province. 
Slag has been discovered and its metal composition is analogous to the ores of 
the Nurata mountains (Bogdanova-Berezovskaya 1962, pi. 12; Kuz'mina 1966: 
91,92). 

In its shape and ornament, the ceramics are analogous to those of Khorezm. 
Ninety percent of the decoration is executed in plain stamp. But the pots with a 
ledge of the Alakul' type outnumber the globular ones with a narrow neck. 

In the five graves of the flat -grave cemetery of Gurdush and the Kyzylkyr bu- 
rial the dead lie in earth pits in the flexed position, head to the west, accompa- 
nied by one or two vessels (Fig. 45: 9, 10), the women wearing bracelets, figure- 
of-eight earrings, beads on the boots, and necklaces of colored stones. The simil- 
arity of the culture -determining characteristics: type of settlement, burial rite 
and, most importantly, the ceramic complex, as well as the cultural and econo- 
mic type makes it possible to include the monuments of the Lower Zeravshan in 
the range of the Tazabagyab culture. The Zeravshan culture took shape as a 
result of the settling to the south of either the western Andronovo tribes or the 
already full-fledged Tazabagyab populations. The newcomers replaced the more 
ancient population of the Zaman-Baba culture. Its participation in the ethnogeny 
manifests itself in the preservation of egg-shaped vessels and small pots with 
inflated sides and a cylindrical neck, in the type of stone laurel-leaf arrows and 
various beads of marble, cornelian, turquoise, and lapis lazuli, including cross- 
shaped ones. The date of the sites is set in the third quarter of the 2nd millennium 
BC based on analogies with Khorezm and the absence in the ceramic complexes 
of applied -roller ware. 

Sites of the Central Zeravshan 

The ethno-cultural map of the Samarkand region is exceptionally diverse. In the 
Muminabad cemetery the Alakul' rite of burying in the flexed position, head to 
the west, comes together with the biconical vessel of the Timber-grave type and 



THE SETTLEMENT OF PASTORAL TRIBES IN CENTRAL ASIA 24 1 

a rich variety of decorations (bracelets, temple rings, rings, numerous beads 
sewn onto clothes and shoes, a mirror with a handle-loop and Fedorovo trumpet- 
shaped earrings (Lev 1966; Askarov 1969, 1970). 

In the Chakka cemetery, the typically Timber-grave burial rite (kurgan, 
flexed position of the dead, head to the north) occurs along with a biconical 
vessel of the Timber-grave type as well as typically Andronovo bracelets with a 
conical spiral and a trumpet-shaped earring (Krikis 1975). Of the same mixed 
character are the burials of Siab (Fig. 85) and Saygus (Dzhurakulov 1986) and 
the ceramics from the site in the Engels kolkhoz and in the overlying layer of the 
Sazagan II settlement of the Neolithic period. At this settlement there were 
discovered hearths, animal bones, an Andronovo trumpet-shaped earring, and 
Fedorovo-type ceramics, which combine with the typologically later pottery of 
the Burgulyuk culture (Dzhurakulov and Avanesova 1984: 32-39). 

The syncretic type of monuments does not allow us to identify stable 
characteristics of ritual and ceramics specific for a particular culture or its 
variant. This provides no grounds for accepting the differentiation by Avanesova 
(1985: 39) of the Zeravshan's monuments of both the Samarkand and Bukhara 
regions, as a single Zeravshan variant of the Andronovo-Tazabagyab commu- 
nity, but enables one to only assert the integration process of different groups of 
the Timber-grave and Andronovo population in the third quarter (particularly at 
its end) of the 2nd millennium BC. 

Analysis of Muminabad's bronzes with its high tin content makes it possible 
to suggest the workings of cassiterite in the Zerabulak mine (Naumov 1972). The 
source of tin for the whole of the Zeravshan was the rich deposit of Karnab, 
where the ancient mines of Karnab, Lapas and Changali were discovered as well 
as a nearby settlement, at which four dwellings of metallurgists with stone 
foundations were revealed as well as hearths analogous to Andronovo examples. 
In addition, stone tools and ceramics were found (Parzinger and Boroffka 2001: 
12-17). 

To the group of sites in the vicinity of Samarkand, the chance findings at the 
settlement of Dzham, south-west of Samarkand, can also be attributed, including 
articles of the BMAC type: two vessels, a flat mirror and a mirror with a pro- 
truding handle and a side bearing a geometric ornament, a pin with a deer at the 
top, a double-edged adze as well as an Andronovo double -bladed arrow with a 
protruding plug; besides, a complex of Andronovo ornaments, possibly origina- 
ting from a burial, included Fedorovo trumpet-shaped earrings, a temple ring 
with open ends, and beads (Avanesova 2001: 63-73, fig. 2-4). These findings bear 
out the existence in Central Zeravshan of the BMAC farmers and Andronovo- 
Fedorovo population (Fig. 84: 1-11). 

The Tashkent Oasis 

Sites from this region are very poorly studied. The oasis, which was located at 
the crossroads of migration routes, was the zone of active contacts. 

The Timber-grave culture 

To the Timber-grave culture belong: the kurgans of Yangiyul with a burial in an 
earthen grave in the flexed position on the right side, head to the east, 



242 CHAPTER NINETEEN 

accompanied with ocher and a late Timber-grave unornamented pot; three 
burials in Saraagach Minvody where the dead lie flexed, head to the north, 
together with unornamented pots of the late Timber-grave type; a kurgan with 
inhumation and vessels on the Nikiforovo lands in Tashkent; the grave of 
Orekhovskoe with a Timber-grave type biconical pot bearing herring-bone 
ornament (Fig. 47; 6) and the possibility of interpreting the semantics of 
Andronovo art and rituals in the light of Indo-Iranian mythology and ritual 
practice Terenozhkin 1940: 91; Oboldueva 1955: fig. 62). Also to the Timber- 
grave circle of sites belong the axe and adze from the Chimbaylyk hoard (Fig. 
114: 17) including a copper ingot. The metal composition is indicative of the 
existence in the oasis of an independent metalworking center which utilised the 
polymetallic deposit of Karamazar (Kuz'mina 1966: 93). The Tashkent oasis was 
the eastern outpost of the distribution of the Timber-grave community. 

Fedorovo type sites of the Andronovo culture 

To the Andronovo culture of the Fedorovo type belong several burials in 
Vrevskaya that contained the charred remains of animal (?) bones and vessels 
including a vase-like pot ornamented with a fine-indented stamp (Fig. 45: 7; 
Voronets 1951: 68; Litvinsky 1962: pi. 109). 

The late Fedorovo population dwelt also in the mountains surrounding the 
Tashkent oasis. On the Chirchik river in Aurakhmat and Iskander inhumation 
burials have been discovered in kurgans with a stone cairn. The women wear 
four to five bracelets with cast conical spirals. They are typical of the early 
Andronovo metallurgical province and date back to no earlier than the 13 
century BC. 

In Brichmulla and Gazalkent there are also kurgans with inhumations 
(Voronets 1948; 1951: 68, 69; Litvinsky 1962: pi. 109; Kuz'mina 1966; 71, 72; 
Rakhimov 1970: 41; Duke 1979: 44, fig. 45). Late Andronovo pots were found 
in Zangiata on the Tashkent -Chimkent road. Information about the other sites of 
the oasis is incomplete and the ceramic assemblage is not distinctive enough to 
establish its cultural origin. These are the sites of Serkali, burials of Khodzhikent 
and Kyzyl-Tu, the discovery of vessels in Angren, Ak-Tepe II, Childukhtaron, 
Brichmulla, Akhangaran (Litvinsky 1962: pi. 109; Maksimova et al. 1968, fig. 8; 
Rakhimov 1970; Duke 1982: 83, 84; Avanesova 1985: 38; Drevnosti Charvaka 
1976: 36, 41, fig. 8: 17). These enable one to include the oasis in the range of 
occupation of the steppe tribes. 

Neither the early Fedorovo, nor the early Timber-grave complexes have been 
revealed so far. The date of the monuments is the late third quarter of the 2nd 
millennium BC. The upper margin is the 1 1th century BC which coincides with 
the spread of the new culture of Burgulyuk (Duke 1969, 1979, 1982a; Drevnosti 
Tuyabuguza: 72-80). 

Andronovo sites of the Fedorovo type in Kirgizia 

Some of the Andronovo tribes changed over to full mobile pastoralism that 
allowed them for the first time to reach the mountain pastures. Arpa, the highest 
Andronovo mountain cemetery, is located in the Tian-Shan mountains at a height 
of 3000m above sea level. A Fedorovo settlement was discovered in Prigorodnoe 



THE SETTLEMENT OF PASTORAL TRIBES IN CENTRAL ASIA 243 

near the town of Bishkek, and a vessel was found at Issyk-Kul' (Bernshtam 
1952: 19, 20; fig. 7; Kozhemyako 1950: 103, 104; fig. 16, 17). 

In Arpa fifteen kurgans were identified with pebble and earthen banks, 12- 
13m in diameter and 0.3-0. 6m in height, covering over round, square or 
rectangular enclosures of stone slabs set on edge. The burial rite was cremation. 
The graves contain 2-3 vessels each; in one there are two bracelets, multi-lobed- 
shaped beads and ribbed paste strings. In Prigorodnoye the above-grave 
construction is not preserved; the dead were buried in a pit, in flexed position on 
the left side, head to the west. In the grave there were found temple rings, a 
vessel, and a shell. 

The stone constructions, the rites of cremation and inhumation, and the 
ornaments are typical of the Fedorovo type sites. The polished pots are decorated 
with the typically Fedorovo ornament in the form of oblique triangles, pennants, 
meanders, and X-shaped figures executed in indented stamp. The vessel from 
Prigorodnoe has the decoration applied over three zones; on the pots from Issyk- 
Kul' the rim and neck zones are united which is a peculiarity of the Andronovo 
pottery of Central Asia, thus allowing one to suggest the possibility of 
differentiating in the future the Tian-Shan type of monuments of the Fedorovo 
line of development. The Tian-Shan sites are dated to the third quarter of the 2nd 
millennium BC on the basis of analogous perforated shell beads and strings from 
Andronovo and Timber-grave sites. 

The genesis of the Tian-Shan sites is related to the Fedorovo sites of central 
Kazakhstan, which was connected with Kirgizia by the Chu and 111 rivers. The 
Tian-Shan complexes provide the earliest evidence of the development of 
pastoralism in the high mountains. 

The penetration of Kirgizia by a group of the Alakul' type is demonstrated by 
the small collection of ceramics from a site near the Belovodskaya fortress 
(Kuz'mina and Mokrynin 1985). In southern Kazakhstan in Semirech'e the 
Kapal cemetery has been discovered whose ceramics display a prevalence of 
Alakul' characteristics (Karabaskokova 1991). 

The Semirech 'e-Fergana type 

This type of monument is represented in Kirgizia, southern Kazakhstan and 
southern Fergana. Therefore it is more appropriate to refer to it as Semirech'e- 
Fergana type. 

Early monuments have been recorded in north-eastern Semirech'e: the 
cemeteries of Talapty I, II, Kuygan II, Mynchunkur III, IV (Karabaskakova 
1987; 1989; 1991). These display a mixture of the Alakul' (vessels with a 
ledge) and the clearly predominant Fedorovo features in the burial rite (cists, 
combination of cremation and inhumation) and ceramics (vessels with rounded 
shoulder). To the Fedorovo type belongs a temple ring with two horses 
analogous to the horses on a knife from Seyma, which indicates the date of 
these sites (Fig. 68: 4). They are genetically related to those of southern 
Kazakhstan. The later Semirech'e proper type is represented by the sites of 
three local groups: Semirech'e, north-eastern Tian-Shan and southern Fergana. 
In southern Kazakhstan cemeteries have been examined at Kara-Kuduk 
Alakul' and Koktuma (Maksimova 1961; Kushaev 1968), in Kirgizia - Tash- 
Tyube II, Tash-Bashat, Besh-Tash, Tegirmensay, Dzhazy-Kechu, Karak-mat, 



244 CHAPTER NINETEEN 

Dzhal-Aryk, Kul'an-say, Kyzyl-say, Chon-Kemin (Fig. 73b; Kozhemyako 
1960; Abetekov 1963; Kuz'mina 1970; Kozhomberdiev and Galochkina 1969; 
1972; Galochkina 1977; 1996; Galochkina and Kozhomberdiev 1995); a large 
collection of ceramics and metalwork has been found on the Bol'shoy 
Chuyskiy Canal (Bernshtam 1950, pi. xlviii: 7-10) and also on the settlements 
of Bishkek Kainda, Dzhal-Aryk (Terenozhkin 1935; Bernshtam 1950: 105, 
fig. xxix; xxx; Kozhomberdiev 1977; Kozhomberdiev and Galochkina 1969; 
1972). 

In Dzhal-Aryk traces of constructions were revealed, made of soman (clay 
with an admixture of straw), large fire sites, ash-pits and ceramics. 

In these cemeteries different types of constructions occur side by side: 
square, round, oval and rectangular stone enclosures (3 x 3 to 5 x 5 m) are 
predominant; occasionally, they are co-joined (Fig. 1: types III A; IV A; VII 
A; VIII A) or have several extensions. In the center of the enclosure there is a 
burial pit (0.7-1.4-1.4-2.2 m 2 ) or a stone cist, and in one case (Alakul') a 
wooden srub (framework). Sometimes the top of the grave is covered with 
stones. The pits are occasionally filled in with crushed stone, the walls and the 
bottom are coated with clay (Besh-Tash). 

In Kazakhstan only inhumations have been revealed so far. In Kirgizia in 
the Kulan-say and Kyzyl-say cemeteries there are also only inhumations, the 
other burial grounds being bi-ritual — with cremation and inhumation co- 
existing within a single enclosure and even the same pit. The dead are 
normally placed in the flexed position on the left side, head to the west. 
Cremations have been identified in 70-80% of the graves. 

The accompanying items for both rites are identical: vessels, bell-mouthed 
earrings, often coated with a gold leaf (Tash-Tyube, Tash-Bashat, Tegir- 
mensay, Kok-Tuma), earrings with a conical spiral (Dzhazy-Kechu), bracelets, 
round plaques, cast and forged beads, pendants, needles, awls, a single-bladed 
knife (Tash-Tyube). 

The late Fedorovo type pots with a rounded shoulder are prevalent, less 
frequent — with a ledge; jars are also present. Most of the vessels are 
unornamented. The ornament (zigzags, notches, in one case — a rhombus) is 
applied to the shoulder, rarely on the neck, in plain or indented stamp. 

At the settlement of Dzhal-Aryk the same pottery combines with rough 
vessels with swollen sides and rims of complex profile, occasionally with 
notches along the edge of the rim and zigzags and herring-bones over the 
shoulder, and in Kainda and places on the Bol'shoy Chuyskiy Canal also with 
vessels with applied-rollers of the Alekseevka and Dongal types. The 
chronology of the sites ranges from the 1 3th to 9th centuries BC, mainly, the 
11th to 9th centuries BC. This has been determined, first, by the discovery of 
Alekseevka and Dongal type ceramics on settlements, a single-bladed knife, 
cast beads; and second, by the filling in and covering of graves with pakhsa, a 
custom that appeared in central Kazakhstan in the early Dandybay period 
(Margulan e? al. 1966: 160-164). 

Scholars unanimously assign the Semirech'e sites to the Andronovo 
culture. I distinguished them as a peculiar type and substantiated their 
assignment to the Final Bronze Age (Kuz'mina 1970: 44-48), which was 
accepted by I. Kozhomberdiev and N. Galochkina (1972: 39; 1977: 12, 36). 



THE SETTLEMENT OF PASTORAL TRIBES IN CENTRAL ASIA 245 

The settlement of Semirech'e would derive from central Kazakhstan. The 
Semirech'e type is likely to have formed as a result of the assimilation of the 
Fedorovo population that arrived here earlier with Alakul' groups. The mixed 
population inherited from their Fedorovo ancestors stone constructions, crema- 
tion, the shape of pots without a ledge and their technique of formation, begin- 
ning with the round bottom, bell-mouthed temple rings. The Alakul' traditions 
manifested themselves in the shape of pots with a ledge and the ornamentation 
in plain stamp. Contacts with central Kazakhstan continued in the Final Bronze 
Age when Semirech'e saw the distribution of ceramics with the applied-roller 
and the custom of filling in and covered graves. The Semirech'e burial grounds 
are most similar to the burial ground of Aydarly in central Kazakhstan. Stone 
cists within square stone enclosures revealed flexed burials with a westward 
orientation of the head. Unornamented ceramics prevail, and vessels with 
applied-roller and the Dandybay type are rarely found (Margulan et ah 1966: 
183-186, pi. xviii; xix; fig. 93, 94). 

In contrast to central Kazakhstan, where in the Final Bronze Age the Begazy 
culture took shape under Karasuk influence, this influence is barely traceable in 
Kirgizia, contrary to the opinion of A. N. Bernshtam (1950: 106), who singled 
out a Karasuk stage there. Only in the Vorontsovka cemetery was there found a 
single-bladed knife with a T-shaped section of the handle and one Karasuk 
vessel; and at the settlement of Dzhail'ma several Karasuk fragments are 
included in the complex of late Andronovo ceramics with applied-roller 
decoration. 

The late Semirech'e monuments constitute the direct genetic basis of the 
Iranian-speaking Saka people, which is borne out by the continuance of 
Andronovo traditions in the Early Iron Age and by the continuity of the 
anthropological type. 

The economy of the Semirech'e tribes was based on mobile pastoralism 
which is demonstrated by the topography of the majority of the cemeteries 
located on mountain slopes and in the high mountains. Seasonal nomadic 
movements predetermined the possibility of further migrations of the Andro- 
novans and a change-over to nomadic stock-breeding by the Saka people. 

The other pivotal branch was metallurgical production. To the Semirech'e 
tribes belong the hoards of southern Kazakhstan: Alekseevka, Kamennoe 
plateau, Turksib (Akishev and Kushaev 1963) and Kirgizia: Sukuluk I, II, 
Sadovoe, Issyk-Kul', Shamshi, Tuyuk Karakol I, II, and numerous chance 
finds (Figs. 43a, b; 74-77; Bernshtam 1941; 1950; 1952; Zima 1948; Kibirov et 
ah 1956; Kuz'mina 1961a; c; 1965c; 1966: 94-98; 1968; Vinnikand Kuz'mina 
1981; Kozhomberdiev and Kuz'mina 1980; Degtyareva 1985). 

The types of articles are diverse: lop-headed axes with a cock-comb butt, 
fluted and wedge-shaped chisels, adzes with a projection, sickle-choppers, 
sickles with a coiled hub and rims, celts, celt -hammers, celt-spades, pick-axes, 
various knives, spears — socketed and shafted, a pole-ax, daggers, double- 
bladed arrows, and ornaments. They are dated to the 13th 9th centuries BC 
based on the co-occurrence in the complexes of the types, which find analogies 
in the Final Bronze Age complexes of the Eurasian steppes (Kuz'mina 1965; 
1966: 94-95). It enables one to establish the existence of the wide cultural ties 
of Semirech'e in the Final Bronze Age. The hoards of Sukuluk II, Karakol II 



246 CHAPTER NINETEEN 

and Issyk-Kul' are typologically differentiated as a later group dating to the 
1 lth-9th centuries BC. 

In terms of typology the metalwork of Semirech'e belongs to the eastern- 
Andronovo metallurgical province (Kuz'mina 1965; 1966: 95) and constitutes 
within its bounds an independent metallurgical seat. It is documented by the 
findings of a casting mold (Aleksandrovskoe), smith's hoards (Sukuluk II, 
Karakol II) and the specific composition of the metal corresponding to the 
polymetallic ores of the deposits of Aktash, Ketmen' -Tyube, those on the river 
Chu, and on the Issyk-Kul', near Alma-Ata. 

The appearance of family hoards, incorporating various types of 
implements, reflects the social stratification of the population of Semirech'e. 
That the hoards were buried suggests that the Final Bronze Age was a period 
of intensified military clashes among the pastoral tribes over ore deposits and 
fertile pastures. These processes were related to the ecological crisis that 
accelerated the change-over to nomadic pastoralism and the development of 
new ecological niches in the mountains, that required vertical nomadic 
movements through the passes. As a result new routes were blazed, in 
particular, into Xinjiang along the Hi river and through the Dzhungar gates, 
along the way of the future Great Silk Road (Kuz'mina 1996; 2001; Sala 
1999). 

Sites of the North-eastern Tian-Shan 

In the mountains of the North-Eastern Tian-Shan there were discovered the 
settlements of Talapty, Turgen' I, II, Uzunbulak and Asy (Fig. 73a). In Asy at 
a height of 2400m above sea level a large house was excavated. The walls of 
this semi-subterranean house were formed from stones and fortified with posts; 
hearths and ceramics were found in the center (Mar'yashev 2001). In the Oy- 
Dzhailyau and Tamgaly ravines cemeteries were excavated at Kul'say I, 
Uzunbulak I (Fig. 73a), Kyzylbulak I, II, Tamgaly I-IV (Mar'yashev and 
Goryachev 1993a, c; 1999; 2001; Rogozhinsky 1999). 

The sites can be assigned to two stages. At Tamgaly rectangular and 
rounded stone enclosures were discovered, occasionally burial mounds formed 
from cobbles (types III A, IV A, VII A). In the center there was a stone cist. 
The burials comprised inhumations, the dead lying flexed on the left side, head 
to the west. Of particular interest are three burials in large clay vessels. An 
analogous rite is described in the Vedic literature of the Indo-Aryans. 
Cenotaphs are numerous. 

In the graves there were found vessels, round temple rings, and beads. The 
ceramic complex is diverse: type I - pots of the Fedorovo type with a rounded 
shoulder and rich ornamentation, executed by oblique netting in plain or 
indented stamp. The decoration (oblique triangles, pennants, X-shaped figures, 
flutes) is applied over three zones; type II is represented by coarse pots with a 
rounded shoulder, type III by jars. The last two types are either devoid of 
decoration or ornamented on the upper part by horizontal zigzags applied in 
plain stamp. One vessel is decorated with an applied-roller. The date of these 
cemeteries is set to the 13th 12th centuries BC on the basis of the Fedorovo 
type temple rings, Fedorovo traditions of pottery-making, along with the 
predominant ceramic complex characteristic already of the Alekseevka type of 



THE SETTLEMENT OF PASTORAL TRIBES IN CENTRAL ASIA 247 

the Final Bronze Age. Typical of this time is also the trapezoid shape of some 
of the cists. 

The second group comprises the cemeteries of Kulsaj, Uzunbulak and 
Kyzylbulak I, II. In them there were revealed rectangular, often interlinking, 
enclosures of stones placed horizontally in 2-3 rows or of vertically set slabs 
(types IV A,B, VII A,B). The graves contain frames made of logs of the Tian- 
Shan spruce or are overlaid with wood. The main burials are cremations, but in 
the peripheral graves, particularly children's, the inhumation rite is customary. 
There are known one single and one double burial of the cremated remains 
placed into large clay unbaked pots (Goryachev 2001: 53, fig. 7: 1,2). 

Analogies to this rite are known in the Tamgaly VI cemetery in the Tian- 
Shan, Shet I in central Kazakhstan, in northern Bactria and northern Pakistan 
(see below). 

In the graves there were found ceramics, bell-mouthed earrings, pendants, 
beads, and plaques with a loop (Goryachev 2001: 56, fig. 9). 

The ceramics comprise pots with a rounded shoulder of the Fedorovo type, 
more rarely are those with a ledge, basins and jars. The pottery is unornamen- 
ted. This ceramic complex is typical of the Final Bronze Age cemeteries of the 
whole of Semirech'e, which permits one to date these burial grounds to the 
1 1 th 9th centuries BC. To the same period is assigned the pottery of the neigh- 
boring settlements (Turgen I and Asy), comparable with the ceramics of the 
Dongal and Trushnikovo types in eastern Kazakhstan (Mar'yashev 2001: 96). 

A. Mar'yashev and A. Goryachev (1993; 1999; Goryachev 2001: 72) sug- 
gest that this group of sites should be singled out as a distinct Kul'say culture; 
they note analogies in the Fedorovo sites of Siberia and central and eastern 
Kazakhstan (Mar'yashev 2001: 96). I deem that these cemeteries constitute just 
a local variant of the Andronovo Semirech'e type, with which they are linked 
by all the culture -determining characteristics: burial rite, ceramics, ornament. 
The origin of these sites appears to be associated mainly with central 
Kazakhstan, since in Siberia and eastern Kazakhstan the inhumation rite was 
predominant. 

The importance of the discovery of northern Tian-Shan sites consists in the 
fact that here we can trace the dynamics of the development of high-altitude 
nomadic pastoralism. It is further essential that here one has established an 
indisputable association of the settlements and cemeteries with the 
petroglyphs, which for the first time provides reliable grounds for dating these 
petroglyphs and linking them with the Andronovo culture (Francfort et al. 
1997; Mar'yashev and Goryachev 1998; Mar'yashev et al. 1998; Rogozhinsky 
2001). 

Sites of southern Fergana 

In southern Fergana the cemeteries of Vuadil', Karamkul', Arsif, Kashkarcha, 
Yapagi, Chek Tashkurgan, Urukzor were discovered; ceramics have been 
found at Shor-tepe, Osh and the Osh region (Gamburg and Gorbunova 
1956;1957; Gorbunova 1972; 1979; 1995; Piotrovsky 1973; Ivankov 1988). 
The chance findings of celts, celt-spades, bracelets with horns, and bell- 
mouthed earrings (Figs. 70: 3,4; 114: 11,12; Zadneprovsky 1962; 1994; 1997; 
Litvinsky 1962; Kuz'mina 1966) are associated with these sites. 



248 CHAPTER NINETEEN 

In the cemeteries different types of constructions are combined: round, 
square and rectangular stone enclosures predominate; occasionally, we find 
linking ones (see Fig. 1: A, IV A; VII A); rare are burial mounds with a bank 
of earth and crushed stone (type II A) surrounded by a stone circle; also known 
are stone cists without an enclosure. The burials were deposited in stone cists, 
sometimes overlaid with slabs. A peculiar feature of the Fergana burial 
grounds are burials in catacombs (Yapagi, Kashkarcha, Uryukzor). All the 
burials follow the inhumation rite, the dead lying flexed, head to the west 
(Karamkul', Yapagi) and in one grave (Vuadil') — to the east. Each grave 
contains a vessel and sometimes ornaments: bell-mouthed earrings (Vuadil'), 
earrings with a conical spiral (Yapagi, Arsif, Kashkarcha), a mirror (Yapagi), a 
bracelet (Kashkarcha). a bracelet of beads (Vuadil'); in Kashkarcha there is 
also a double-bladed knife and an awl, in Vuadil' — an arrow. 

The vessels have a rounded shoulder and a short neck. Up to 50% of them 
are ornamented, the ornament being applied over the shoulder by carving or in 
plain, very seldom, indented stamp in the shape of zigzags, herring-bones, 
triangles and imprints. This ceramic is a late derivative of Fedorovo forms. In 
Vuadil' and Kashkarcha there were also found vessels with applied-roller 
ornament, characteristic of the Final Bronze Age. In Tashkurgan pottery of the 
Chust culture was discovered. 

The sites are dated to the Final Bronze Age, primarily, 10th-9th centuries 
BC on the basis of the ceramic complex, the button from Kashkarcha and the 
double-bladed arrow with a hidden socket from Vuadil'. B. Gamburg and N. 
Gorbunova (1957) attributed these cemeteries to the late Andronovo culture, B. 
Litvinsky (1962: 287-289; 1983: 157) to the Kayrak-Kum culture, N. 
Gorbunova (1995: 23) to the Kayrak-Kum variant of the Andronovo culture. I 
have emphasized their closeness to the Semirech'e type sites of Kirgizia (1970: 
45). Their origin is likely to be associated with the southward movement of 
late Andronovo tribes from central Kazakhstan or Semirech'e. N. Gorbunova 
(1995: 23), when analyzing the burial rite in the catacombs, specific to 
Fergana, emphasizes the possibility of a southern component in the ethnogeny. 

The culture of Fergana of the Final Bronze Age probably formed as a result 
of an amalgamation of the late Andronovo population with representatives of 
the Chust culture. Most of the cemeteries are located on the mountain slopes, 
which points to a semi-nomadic lifestyle of Fergana's pastoralists. 

Fergana Valley: Late sites of the Kayrak-Kum Type 

On the bank of the Syr-Darya in the Kayrak-Kum desert, seventy settlements and 
industrial complexes have been discovered as well as the cemeteries of Khodzhi- 
Yagona, Dakhana, and Dashti-Asht (Litvinsky 1960, 1962, 1963; Saltovskaya 
1978: 95, 96). Some of the settlements belong to the early Iron Age. The Bronze 
Age complexes have been discovered at settlements 6, 12, 16, and 35. To the 
Kayrak-Kum type also belongs the Ak-Tangi cave and the ceramic finds from 
the Bol'shoy Fergana Canal and nearby town of Suleyman-Tau, and numerous 
examples of metalwork: celt-spades, trumpet-shaped earrings, bracelets with 
conical spirals (Zadneprovsky 1962: 51; 1999; Litvinsky 1962; Kuz'mina 1966: 
94). Generally, the settlements are not large: 0.1-0.3ha, but there are some of 
3ha. All of them are half-eroded. The dwelling type has not been identified. 



THE SETTLEMENT OF PASTORAL TRIBES IN CENTRAL ASIA 249 

Series of stone hearths have been revealed which makes it possible to reconstruct 
the typical Andronovo dwelling up to 20m in length. Outside the settlements' 
boundaries production complexes have been discovered where agglomerations of 
ore and slag have been found weighing up to 1.5-2 tons. At settlement 16, molds 
for a pick were discovered and an axe with a cock-comb (Figs. 74: 3; 113: 28). 
The staple of the economic and cultural type was metallurgy and metal working. 
Judging by the scale of work metallurgy was treated as a specialized craft. The 
copper sources were the Naukat and Karamazar deposits. The Fergana metallur- 
gical center belonged to the eastern Androvovo metallurgical province, charac- 
teristic of which are lop-headed axes with a cock-comb, picks, celt-spades, arrow 
of proto-Saka type, trumpet-shaped earrings, bracelets with concial spirals 
(Kuz'mina 1966: 94). 

The second component of the economy was cattle-breeding. There have been 
found the bones of cattle, ovicaprids and the horse. Indicative of hunting are the 
bones of Gazellas subgutturosa; a fishhook attests the presence of fishing. 

The ceramics of the Kayrak-Kum type are hand-made, occasionally executed 
in a textile pattern. The pots have a rounded shoulder and a short inverted rim, 
often of a complex profile. Apart from this, there occur vessels with globular 
bodies. On late sites one also finds vessels with horizontal handles and kettles. 
Ten percent of the pottery is decorated, half of the ornament being executed in 
combed stamp. The ornament is located in one zone under the rim, rarely also on 
the shoulder. Primitive patterns such as zigzags, herring-bones, and nail impress- 
ions are predominant. 

In the cemetery of Khodzhi-Yagona, stone cists were found built from 
edgewise -placed slabs, oriented northwards. Inhumations are accompanied with 
vessels analogous to those from the settlement, a bracelet, a plate, and beads 
(Litvinsky 1962: 117-118, pi. 55). In the Dakhana cemetery the burials are made 
in stone cists formed from horizontally-laid slabs. The cists protrude above the 
surface and are oriented to the west and south-west. In inhumation burials were 
found one vessel in each, a bronze trumpet-shaped earring, and a stone button 
(Litvinsky 1960; 1962: 158-164, fig. 39-40). The Dashti Asht cemetery dates to 
the end of the period. At ground level under the stone cairns, 4- 12m in diameter 
and 0.4-1. 5m in height, stands a stone cist built from slabs set edgewise or flat 
containing a burial with the dead lying flat on their back head to the west. In the 
burials vessels were found of the Kayrak-Kum and Chust culture, an earring, and 
a plate (Saltovskaya 1978: 95-6; Litvinsky 1962: 246, 248; 1963: 109, 121, 124). 

B. A. Litvinsky (1962: 246, 248; 1963: 109, 121, 124) designated all the 
Fergana monuments as belonging to the Kayrak-Kum culture, assigning to it also 
the complexes of the Tashkent oasis and the Makhan-Darya. This attribution 
appears imprecise (Itina 1967: 78; 1977: 233; Kuz'mina 1966: 94; 1968: 308). 
Later (1983: 157) he only assigned the Fergana sites to the Kayrak-Kum culture. 
N. Gorbunova (1995: 23) assumes that all the Fergana sites, including the 
southern Fergana cemeteries (see below), can be united into the Kayrak-Kum 
variant of the Andronovo culture. The diagnostically important characteristics - 
large houses, burials in stone cists oriented westwards, technology of ceramic 
manufacture employing a textile pattern, shapes and ornaments of the 
ceramics — all of these are Andronovo characteristics. We thus may single out 
the Kayrak-Kum sites as a special type of the Andronovo community. The 



250 CHAPTER NINETEEN 

existing differences in both ritual and ceramics, from my point of view, reflect 
not cultural but chronological peculiarities. The early Andronovo characteristics 
are absent in the Kayrak-Kum ceramics. The dating of the sites to the 12th— 9th 
centuries BC is determined on the basis of the cock-combed axe and the cera- 
mics with a complex-profiled rim. A date of the early 1st millennium BC for the 
Dakhana cemetery is based on the find of a block-like button and for Dashti-Asht 
and Takyri-Yagona on the evidence of vessels from the Chust culture. Kayrak- 
Kum ceramics have also been found on the Chust settlement of Tashkurgan. The 
genesis of the Kayrak-Kum-type sites is likely to be associated with migrations 
of the eastern Andronovo population, which is indicated by the burial rite and the 
eastern Andronovo types of metal articles: cock-combed axes, and trumpet- 
shaped earrings. Particularly instructive is the similarity with the sites of the 
Semirech'e type. 



CHAPTER TWENTY 

RELATION OF THE ANDRONOVANS WITH THE POPULATION 
OF XINJIANG AND CENTRAL ASIA 

It is reasonable to presume that Andronovo influence extended as far as China. 
In the Anyang culture we find the momentous achievements of a world 
civilization — metallurgy, wheeled transport and horse-breeding — already in their 
developed form; the Yellow River displays no preceding development. Pursuant 
to the ancient tradition created during the formation of the Chinese state, 
civilization emerged there independently. This traditional hypothesis of 
autochthonous development has been embraced by most Chinese archaeologists 
(Cheng Te-K'un 1961; Chang Kwang-Chih 1959; 1965; 1968). 

In accordance with another hypothesis put forward by M. Loehr (1949; 1957; 
1965) and the outstanding Russian scholar S. V. Kiselev (1960) (and accepted by 
Li Chi (1957), W. Watson (1961), E. Kuz'mina (1973), Ping Ti Ho (1975), S. 
Kuchera (1977), M. Kryukov, V. Safronov, N. Cheboksarov (1970), and A. 
Varenov (1983)), the formation of Chinese civilization was stimulated by a 
western impulse. In the Eurasian steppes metallurgy, wheeled transport and 
horse-breeding go back to the 4th millennium. BC, while the types of celts, 
spears and single-edged knives of Anyang find their prototypes and analogies in 
the Andronovo and Seyma-Turbino complexes. 

Now it has been established that metal appeared in China in the pre -Anyang 
period on the northern periphery in the cultures of the 'significant others' 
(Linduff 1996b), people, ethnically non-Chinese (Prusek 1971; Wu En 1985; Lin 
Yiin 1986). These cultures have been systematized by K. Linduff (1994; 1995; 
1996a; 1996b; 1997; 1998). The Qijia culture in Gansu (2500-1900 BC) has 
yielded the oldest barley and wheat, horse, forged copper awls and rarely cast 
bronze awls, knives, celts, gold rings, a mirror, plaques, and earrings (Debaine- 
Francfort 1995: 320, fig. 19, 61). In the cultures of Zhukaigou (phases 3, 4, 5) in 
Inner Mongolia (2000-1500 BC), Lower Xiajiadian in the north-east of Inner 
Mongolia and Hebei (2000-1600 BC); Erlitou (periods 3,4) in the Central Plain 
(1750-1530 BC), (Chang Kwang-chih 1968) and Yueshi in Shandong (2000- 
1600 BC) there are incipient signs of the productive economy (pig, horse) and 
metallurgy. In China alongside metal there appeared wheat, barley and sheep, all 
cultivated in the Near East and diffused in the 3rd millennium BC into the 
steppes; the horse was domesticated in the steppes. This testifies to a north- 
western impulse. The multi-ethnic population of northern China apparently 
played a pivotal role in the spread of the productive economy, horse-breeding 
and metallurgy, into the Central Plain from the north, the steppes (Linduff 1994; 
1995a; 1997; 1998; Fitzgerald-Huber 1995; 1997). 

Relations with the north may have been realized via Xinjiang and along the 
steppe corridor of Gansu (Map 15). In the north Xinjiang is connected with 



252 CHAPTER TWENTY 

Siberia by a pass through the Altai mountains. In the west it is linked with Fergana 
by the Tersek Davan Pass and by an easily passable route along the Hi river 
through the Tian-Shan with Semirech'e. The ecological conditions of eas-tern 
Turkestan are very diverse: from the north it is circumscribed by the Altai 
mountains, from the south the Pamirs, Kunlun and Altyn mountains, from west to 
east the Tian-Shan separates Dzungaria from the Tarim Basin; most of the territory 
is occupied by the Taklamakan Desert but in places fertile river and lake valleys 
are suitable for farming, while areas of steppe may be used for stock-breeding 
(Petrov 1966; 1967). This determined the diversified character of the economic and 
cultural types of the region. The Afanas'evo culture was the first in Xinjiang with a 
productive economy. It is represented by a cemetery near Urumchi and Ke'ermuqi 
(Keremchi) in the Altai district (Wang Binghua 1996: 75; Molodin and Alkin 
1997). In terms of the funeral rite and implements similar to the Afanas'evo culture 
we have the Gumugou (Qawrighul) cemetery; the absence of pottery makes it 
impossible to assign an exact attribution (Debaine-Francfort 1998; Mair 1995; 
Mallory 1995; Mallory and Mair 2000; Kuz'mina 1998; Khudyakov, Komissarov 
2002: 31-33). The calibrated date of Gumugou is 2030-1815 BC. The population 
raised cereals, sheep, goats, cows and Bactrian camels, produced textiles of a 
European type (Barber 1998), wore the traditional dress of the steppe-dweller: a 
cap, caftan, trousers and boots, and used forged copper articles. The population 
belonged to the Caucasoid anthropological type (Alekseev 1988; Han Kangxin 
1994; 1998; Chikisheva 1994). The Afanas'evo culture is genetically related with 
the Pit-grave and, partially, Catacomb cultures (Kiselev 1949; Vadetskaya 1986; 
Tsyb 1984; Novgorodova 1989). The arrival of representatives of the Afanas'evo 
culture in Siberia, Tuva and Mongolia is viewed as the first wave of the migration 
of the Indo-European-Tocharians eastward (Semenov 1993), the creators of 
Gumugou being also numbered among them (Jettmar 1985; Mallory 1995; 1998; 
Mallory and Mair 2000; Pulleyblank 1996; Renfrew 1998; Kuz'mina 1998). 

Northern Chinese populations may have received metal, wheat and barley, 
wheeled vehicles, the sheep and the horse from the Afanas'evo tribes, who came 
from the west. The words for all these were borrowed into Chinese from Indo- 
European, presumably Tocharian (Pulleyblank 1996: 1-24). It is likely that the 
rites of domestic animal sacrifice, familiar in the European steppes from the 4th 
millennium BC, were also adopted. In Siberia the Afanas'evo culture was 
succeeded by the Andronovo culture of the Fedorovo type, which came from 
eastern Kazakhstan. In the Fedorovo burial grounds reminiscences of the 
Afanas'evo ceramic tradition can be distinctly traced, but genetically these 
ethnoses differ. As already mentioned, at the early stage of Novy Kumak Andro- 
novo tribes organized large-scale metallurgical production. The history of 
Andronovo metallurgy is closely associated with that of Turbino-Seyma, studied 
by E. Chernykh and S. Kuz'minykh (1989). Turbino-Seyma bronzes are an 
assortment of types, comprising celts, adzes, double-edged knives-daggers, 
single-edged knives, often with a figured handle, spears, including those with a 
socketed shaft, hooks and bracelets. Turbino-Seyma bronzes are distinguished by 
the use of tin bronze and the casting of thin-walled celts, chisels and socketed 
spears with an all-metal socket. The invention of a strong bronze alloy, which 
enabled the production of implements with a cast socket, was a momentous 



RELATION OF THE ANDRONOVANS WITH XINJIANG AND CENTRAL ASIA 253 

innovation for the period. The abundant cassiterite deposits of eastern Kazakh- 
stan provided the source of tin. 

Turbino-Seyma bronzes form part of the complexes of entirely different 
cultures united by a system of rivers. From eastern Kazakhstan, where a great 
number of chance finds are concentrated, tin and bronze came by the river Irtysh 
eastward to the Altai along the Ob (Elunino, Tsygankova Sopka) and to the north 
of the Altai along the Om, the Irtysh's tributary (Rostovka, Sopka 2, Omskiy 
Klad), and also by the Irtysh to the north where the Ob and the Irtysh converge 
with the basin of the Ural rivers. From there metal would find its way to the 
Kama (Turbino) and further to the Volga (Seyma). This 'tin' road preceded the 
Great Silk Road which would connect Asia and Europe. 

What is the origin of Turbino-Seyma bronzes? E. Chernykh and S. Kuz'mi- 
nykh (1989: 259-261; 270) dated them to the 16th century BC and presumed that 
the complex formed in Siberia as a result of the interaction between the culture 
of the hunter-fishers of Baikal and that of the Altai's horse-breeders and metal- 
lurgists. A clan of armed nomad-metallurgists would carry out distant raids on 
horseback and spread their products in the west. V. Bochkarev (1986) estab- 
lished the wide European relations of the bronzes and determined the chronolo- 
gical sequence of the complexes: Turbino-Seyma-Rostovka and Samus' IV, 
Turbino being synchronized with the Abashevo culture. O. Kuz'mina (2000: 65- 
134) confirmed the relationship between the metallurgy of the European Aba- 
shevo culture and Turbino and demonstrated that many types of early Andronovo 
metal from Sintashta advanced traditions of Abashevo metalworking (adzes, 
double-edged and single-edged knives, spears, shafted arrows, hooks, bracelets). 
This bears out the role of the European traditions for the metalworking of 
Turbino-Abashevo-Sintashta. However, there is a group of bronzes of eastern 
origin at Seyma (Chernykh 1970: 155-173). In Sintashta two articles of tin 
bronze alloyed with lead (Pb) and antimony (Sb) were found, which points to 
their being exported from eastern Kazakhstan (Zaykova 2000). That is also the 
likely source for the single -edged knife with a representation of an argali known 
from Turbino and one from Seyma's one with a pair of horses (Bader 1964: 115- 
123, fig. 113; 1970: fig. 52). 

This allows one to pose a working hypothesis concerning the formation of 
Turbino-Seyma bronzes as a result of the interaction between the population of 
Eastern Europe (above all, Abashevo and, partially, Catacomb tribes) and early 
Andronovo tribes of the Fedorovo type in eastern Kazakhstan. There bronzes 
have already been recovered from the Kanay cemetery which preserve the 
Eneolithic traditions, and the early Marinino stage of the Fedorovo type has been 
established (Tkacheva 1997: 12). The acceptance of the calibrated C 14 dates of 
the Abashevo (Mikhaylova and Kuz'mina 1999: 119) and early Andronovo sites 
of the Novy Kumak stage compels one to assign Turbino to an earlier time and 
date it to the turn of the 3rd-2nd millennia BC. It may be presumed that it is the 
early Andronovo tribes of Siberia and eastern Kazakhstan that were instrumen- 
tal in the appearance in China, in the northern contact zone, of tin and bronze 
articles, stone and clay molds, the technique of casting celts and spears with a 
concealed socket, as well as types of adze, the single-edged knife and the ring- 
headed dagger in Erlitou in Henan (Linduff 1994: fig. 3,18) and the dagger from 
Zhukaigou, phase 5 (Linduff 1997: fig. 6 bottom). The type of twisted-butt 



254 CHAPTER TWENTY 

daggers with animal figures on the handle was widely developed in China (Lin 
Yiin 1986: fig. 49, 17; Linduff 1996a: fig. 9; Chzhun Suk Be 2000: fig. 2, 1-6). 
China's socketed spears are analogous with Andronovo spears, which go back to 
Abashevo-Turbino prototypes (Loehr 1956; Varenov 1987; 1989). The pitch- 
fork-shaped spear with a hook from Shenna, Qinghai (Wagner 2001) resembles 
the spears from Rostovka and from the Altai (Fig. 109: 15, 16; Chernykh and 
Kuz'minykh 1989: fig. 29, 30). 

Having received this initial western impulse, Chinese metallurgists began to 
develop their own production. On the Yenisey in the 14th 13th centuries BC 
Andronovo-Fedorovo tribes were ousted by the newly arrived tribes of the 
Karasuk culture. In eastern Kazakhstan and Semirech'e the development of the 
Andronovo culture was still under way. In the 13 th " 12 th centuries BC pottery with 
applied-roller and many types of metal articles, common from the Danube to the 
Altai, spread here, and the activity of the metallurgical centers of Semirech'e and 
Fergana intensified. Relations with Xinjiang stepped up. Xinjiang's sites are 
diverse and include: agricultural settlements, stock-breeders' sites, cemeteries, 
hoards and chance finds (Jettmar 1985; 1992; 1996; Molodin and Alkin 1987 
Antonova 1988; Debaine-Francfort 1988; Kuchera 1988; Khavrin 1992 
Kuz'mina 1996a, b; 1998; 2000; Zadneprovsky 1992; 1993; 1994; 1995; 1997 
Semenov 1993; Molodin and Alkin 1997; Molodin 1998; Shui Tao 1998; Ke 
Peng 1998; Mei and Shell 1998; 1999; Mallory and Mair 2000). Of particular 
importance are the works of Debaine-Francfort and Mei and Shell. Xinjiang's 
population belonged to different anthropological types, including Pamir-Fergana 
(Andronovo) (Han Kangxin 1998). The monuments and pottery are diverse and 
attempts to provide them with a local origin and chronological classification 
(Chen and Hiebert 1995; An Zhimin 1998) are so far unconvincing (Komissarov 
1997). There are Andronovo sites in Xinjiang: in the Sazi cemetery in Tuoli on 
the border with eastern Kazakhstan burial mounds composed of stone and an 
earthen bank were discovered and a Fedorovo vessel was found (Mei and Shell 
1999: 573, fig. 3: 1); in the cemetery and settlement of Tacheng (Chochak Tar- 
baghatay District) Andronovo pottery with stamped geometrical ornament in the 
shape of a herring-bone, triangles and zigzags was recovered. A large jug deco- 
rated with herring-bone and a small jug ornamented with nail impressions (Mu 
Shing In 1996: 27, fig. 14), analogous to the vessels of eastern Kazakhstan, have 
been published (Fig. 109: 17, 18). An adze, a pair of earrings, beads and a copper 
ingot were also found there. At the agricultural settlement of Xintala in the 
Tarim Basin (the radiocarbon date is 1700-1300 BC) two archaeological layers 
were discovered. In the lower one painted pottery was discovered together with 
vessels with comb-shaped geometrical ornament, a stone mold for casting an 
awl, and an awl and a knife; and on the surface a celt and a socketed arrow were 
found (Debaine-Francfort 1998: 16; Mei and Shell 1998: fig. 3: 1; 1999: fig. 7). 

Metal articles of the Andronovo type were discovered at other sites in 
Xinjiang (Figs. 54; 74: 7; 75: 1, 13, 17, 21; 76: 8; 77: 4, 10, 13, 15, 16; 109: 4-7, 
11-15). In the Qizilchoqa cemetery of the Wupu group (Hami) (C 14 date is 1350- 
1000 BC) a chisel, a socketed arrow, a mirror with a handle and beads sewn on 
the boots were found (Debaine-Francfort 1988: 18-19, II: 5). In Yanbulaq 
(Qumul group) (C 14 date is 1110-525 BC) 76 burials were excavated and 94 
bronze articles were recovered: single-edged knives, socketed arrows, awls, 



RELATION OF THE ANDRONOVANS WITH XINJIANG AND CENTRAL ASIA 255 

rounded plaques with punched ornament (flat and with an eyelet). A celt and a 
ring-headed knife were found at the Lanzhouwan settlement of the Nanwan 
group (C 14 date is 1385 ± 75 BC). A celt, knives, an arrow, an awl, a mirror with 
a central projection, earrings and beads were excavated from the Nanwan grave 
(C 14 date is 1050 BC). At the Qaraqocho settlement in Turfan (C 14 date is 945- 
100 BC) a sickle and awls were discovered. 

Chance finds are also known in eastern Turkestan. A celt and an arrow or 
javelin were discovered in Kroran by Sven Hedin (Bergman 1935: table xvi 1, 
7), an asymmetrical celt was found in Xinjiang, a celt-spade and a flanged adze 
in Urumchi (Debaine -Franc fort 1988: figures 9, 3, 5). In the area of Tacheng and 
Tuoli (Mei and Shell 1999: 573, fig. 4) two axes, four sickles, an adze, a celt, a 
celt-spade, a chisel and a spear were recovered, part of the articles being made of 
bronze containing 2-10% tin. In the Tian-Shan near Yili an adze was discovered; 
in Yining — a chopper-sickle; in Nileke — a celt; in Xinguan — a chisel; in 
Jimusa'er and Qitai two axes were found; in Hami — a celt and a sickle; in Kuysu 
in Balikun — another ax; in Chaqimale in Huayuan near Hami — an arrow and two 
knives: ring-headed and with a deer's head. 

The hoard discovered in Agarshin in Toquztara is very interesting (Fig. 75). 
It was discovered in 1975 at a depth of 1 metre (Debaine-Francfort 1989: 200, 
fig. 20, table 11, 5, 6; Ke Peng 1998: fig. 1-6; Wang and Cheng 1989: 95, 96). 
The complex was found in the vicinity of graves with stone slabs. It contained a 
red-fired vessel and 13 [12] bronze articles: three axes with drooping butt -ends, 
three sickles, five chisels and a celt-hammer. An Zhimin (1998, photo 2-5) 
includes three more adzes. Originally, the hoard was attributed to the Warring 
States period, to the developed Iron Age (Wang Binghua 1989: 200). C. 
Debaine-Francfort (1989: 200) pointed out the possibility of synchronizing it 
with the Andronovo culture, but she attributed it to the Saka period. Ke Peng 
(1998: 580) assigned it to Andronovo and dated these finds to 1500-1000 BC. I 
set the hoard to the Andronovo culture and dated it to the 13th— 1 1th centuries 
BC (Kuz'mina 1994: 241). The comparison of Xinjiang's bronze artifacts with 
those of Andronovo permits us to specify their chronology and origin 
(Khudyakov and Komissarov 2002: 34-35). 

In the cemeteries of Wubao and Lafuqiaoke (1165-890 BC) bronze single- 
edged knives and ornaments were found (Wang Binghua 1996: 77). 

Thus, metallurgy in China emerged as early as the turn of the 3rd-2nd 
millennia BC under the influence of the Eurasian steppes. It was mediated not by 
the ethnically Chinese tribes of China's northern periphery ('significant others' 
according to K. Linduff 1996b), but, initially, by the tribes of the Afanas'evo 
culture and then the Turbino-Seyma and Andronovo. Borrowed were the 
technology of making a bronze alloy, the use of gold and the casting of spears 
and celts with a concealed socket in two-part molds. Particularly active were the 
relations between Semirech'e, Fergana and eastern Kazakhstan and Xinjiang, 
where an Andronovo population settled and all the specific types of the imple- 
ments of the Semirech'e metallurgical center were in general use. 

Other innovations of Chinese civilization were horse-breeding and wheeled 
transport. Horse bones are represented in the metal-using cultures of the early 
2nd millennium BC (Qijia, Siba and Longshan), but their role was negligible and 
there is no evidence of domestication (Linduff 2000a, b). Judging by the depic- 



256 CHAPTER TWENTY 

tions (Linduff 2000b: fig. 1, 2) and, particularly, engravings denoting the horse 
in Chinese oracle -bone inscriptions, it was the wild Przhevalsky horse (Mair 
1998). Its range in the historical period embraced the whole of the Eurasian 
steppes, including Mongolia. The Przhevalsky horse is untamable. Since the 
number of chromosomes of the domestic and Przhevalsky horses is different, the 
latter could not be the ancestor of the former. The domestic horse may have 
originated from the tarpan in the Pontic-Caspian steppes, where already in the 
4th-3rd millennia BC its cult emerged and the horse was represented in art and 
ritual sacrifices. In southern Siberia the domestic horse is familiar to the 
Afanas'evo culture. But initially the domesticated horse would only be used as 
food (Bokonyi 1995; Kuz'mina 2000; Linduff 2000a). 

The Afanas'evo culture was probably also familiar with wagons, which had 
solid wheels made up of three parts assembled by mortise and tenon and with a 
protruding hub. Draught animals were a pair of bulls or oxen (Leontiev 1980: 65; 
Vadetskaya 1986; Gryaznov 1999). The similarity in construction of the Old 
World's carts and wheels in the late 4th - mid 3rd millennia BC gave rise to the 
hypothesis of their monocentric origin in the Near East (Childe 1954; Piggott 
1969; 1983; Littauer and Crouwel 1979). This type of transport is also familiar to 
the Andronovo culture. For now the appearance of the cart in China is evidenced 
by its representations in petroglyphs in Xinjiang and Inner Mongolia (Kezier- 
queqia (Qizilchoqa) near Qumul in Xinjiang, dating back to 1350-1000 BC 
(Mallory and Mair 2000: 142, 143, 324, 325: fig. 64)). Horse bones have been 
found in Xinjiang on sites of the latter half of the 2nd - early 1st millennium BC: 
Shirenzi, Lanzhouwan, Nanwan, Wupu, Kezierqueqia (Qizilchoqa) (Debaine- 
Francfort 1988: 18-21; Jettmar 1992), and cheek-pieces of steppe type were 
found in Shirenzi and Nanwan. 

The most important innovation of the first quarter of the 2nd millennium BC 
was the spread of the light war -chariot with two spoked wheels, harnessed to a 
pair of horses. The oldest finds of chariots and horses in warriors' graves are 
known from the Urals and on the Volga (Gening 1979; Kuz'mina 1994; 2001; 
representations are known from Anatolia and Syria (Littauer and Crouwel 1979; 
Moorey 1986). Having emerged in the formative period of the Andronovo 
culture at the sites of Sintashta and Petrovka, chariots dominated the steppes in 
the third-quarter of the 2nd millennium BC in the Timber-grave and Andronovo 
cultures, which is evidenced by cheek-pieces, representations on vessels and on 
petroglyphs. The representations of chariots in Xinjiang (Cheremisin and 
Borisova 1999: 129-134, pi. i; ii) are analogous to the Andronovo chariots of 
Kazakhstan and Semirech'e and are executed not in the Near Eastern manner in 
profile but in the Eurasian style en face (Littauer 1977; Novozhenov 1994), 
which indisputably corroborates their origin in the north-western steppes (Fig. 
93: 10-13). 

It is interesting that the graph denoting the chariot in the oracle-bone 
inscriptions (Shaughnessy 1988: fig. 4) resembles the pattern in petroglyphs of 
Central Asia (Fig. 54: 13, 18; Novgorodova 1981). Chariots proper were 
discovered in the 1930s at the imperial cemetery of the Shang dynasty and near 
the palace in the capital of the Yin kingdom in Anyang and later in its 
neighborhood near Beijing, in Xiaotun, Dasikong, Baijiafen, Xibeigang, and also 
in Liulihe, etc. (Kaogu Xuebao 1947 #2; 1955 #5; 1979 #1; Kaogu 1961 #2; 



RELATION OF THE ANDRONOVANS WITH XINJIANG AND CENTRAL ASIA 257 

1972 #4; 1977 #1; 1987 ##5, 12; 1998 # 9; Wenwu 1977 #5; Cheng-Te-K'un 
1960: 71, 260, pi. xviii c; xxv d, map 1; Watson 1961: 64, pi. 48; Dewall 1964: 
124-127; Ping-ti Ho 1975: 354-357; Li Chi 1977: 113-115, fig. 19; Kuchera 
1977: 133-140, 173, fig. 64-67; Kozhin 1977; Kryukov et al. 1978; Varenov 
1980: 164-169; Shaughnessy 1988: 191-194; Linduff 2000a). Beside the graves 
of the kings and elite there were discovered the chemaken pits (literally - 'a pit 
with a chariot and horses'). Their date is 1250-1100 BC. They contained wea- 
pons, a chariot whose wheels were placed into segmented grooves analogous to 
those of Sintashta, and two horses, laid, as in Sintashta, on their side parallel to 
one another (Fig. 112: 1). A quadriga was found but once (Xiaotun M 20). Char- 
iots with four horses were typical of the following Zhou dynasty (Komissarov 
1980; Shaughnessy 1988; Linduff 2000a). Sometimes a charioteer or a groom 
was buried in a chemakeng. 

Cheek-pieces are rectangular, with a central orifice (Fig. Ill: 24,25), made 
of bronze (a Chinese innovation). The harness, as in Andronovo, has a nose- 
strap. In contrast to the Near Eastern wheels, the Chinese ones are multi-spoked, 
like those of Andronovo. Other peculiarities of construction of the Chinese 
chariots are also close to Andronovo, as far as one may judge from the 
petroglyphs. This points to the steppe origin of chariots in China (Dewall 1964 
Kozhin 1969; 1977; 1988; Li Chi 1977; Kuz'mina 1973a; 1977; Piggott 1978 
1983; Varenov 1980; Komissarov 1980; Shaughnessy 1988; Linduff 2000b: 
Mallory and Mair 2000). The northern tribes served as mediators in their transfer 
to Anyang. This is borne out by the ritual oracle -bone inscriptions, which record 
the seizure by the Shang army of rich booty in northern and western China: 
chariots, horses and weapons (Ping-ti Ho 1975: 225; 356-357; Shaughnessy 
1988: 214; 233; Linduff 2000b). 

Apparently, together with the horse and the chariot Yin China also adopted 
the art of horse training, their name, and religious and mythological concepts 
associated with them. The word 'horse' ma is an old Eurasian migrational term 
(Polivanov 1968; Pulleyblank 1966: 11, 12), and the name of the chariot stems 
from the Proto-Indo-European 'wheel' and came to Shang China via either 
Tocharian or early Iranian (Pulleyblank 1966: 30; Lubotsky 1998; Bauer 1994; 
Mallory and Mair 2000: 126). The cult of the chariot and the horse and the rite of 
its sacrifice, particularly at the funeral of a king or military elite, is characteristic 
of the Indo-Iranians, and archaeologically it is attested in the Andronovo culture. 
Chinese myths about the connection of the emperor with the winged heavenly 
horses, which rendered him immortal, the horse coming out of water, the thunder 
chariot and the sun chariot used by the solar god for travelling over the earth, 
arise from the Indo-European and, particularly, Indo-Iranian mythology 
(Bussagli 1955: 17-22; Waley 1955; Dewall 1964: 121; Yuan Ke 1965: 176-177; 
Pulleyblank 1966: 32; Kuz'mina 1974: 83, 84; 1977: 45), and the names of 
fanciful horse-griffins were adopted from the Tocharians or, more likely, Indo- 
Iranians (Izushi 1930: 346-387; Waley 1955), while the images themselves were 
incarnated by the Iranian peoples in Scythian animal-style art, and in folklore 
they have survived till the present time (Kuz'mina 1977). 

Thus, the analysis of the relations between the Andronovo culture and Shang 
China enables the following conclusions to be drawn: 1) cultural relations bet- 
ween the steppes and the Central Plain were established in the 2nd millennium 



258 CHAPTER TWENTY 

BC; at that time sections of the future Great Silk Road (Map 16) were 
established, along which metal, the horse and the chariot reached the Celestial 
empire; 2) since cultural borrowings were reflected in the Chinese language in 
the words adopted from Indo-European, first and foremost, Indo-Iranian, this 
serves as an important — and independent — argument in favor of recognizing the 
Andronovans as Indo-Iranians. 3) the established relations permits one to refine 
the Andronovo chronology by synchronizing it with the Chinese. 

The lower date of the Andronovo type axes is determined by the discovery of 
a bronze model of an axe with an oval socket, reinforced with a raised-border 
evolving into a comb, in a cremation burial in the Stary Tartas IV cemetery 
belonging to the Fedorovo Yenisey type (Molodin et al. 1998: 294-299, fig. 2b). 
The complex of the grave with a richly ornamented square Fedorovo vessel, 
comparable with the Alakul', and horse bones fixes the date of the axe as no later 
than the 14th century BC. 

Of particular interest is the tomb of the ruler Fu Hao, who was a consort of 
the king Wu Ding (c. 1200 BC). The tomb is situated in Xiaotun at the imperial 
cemetery in Anyang (Henan Chutu 1981: 147, 148: fig. 136-181; Linduff 1994: 
418; So and Bunker 1995: 36). Alongside a set of Chinese articles, including 
those with inscriptions, the complex comprises a twisted-butt knife, a bronze 
mirror of the Andronovo type, jade rings and bracelets resembling Seyma types, 
and stone figurines of people and horses, comparable with Andronovo plastic art. 
These articles are probably of northern origin. In the complexes of Anyang two 
types of items can be distinguished: 1) local articles, including those with the 
emperors' names; 2) imported pieces of northern origin: Andronovo temple-rings 
of the Fedorovo type made of gold alien to China, socketed two-bladed arrows 
and spears atypical of China, and single-edged knives with a zoomorphic handle. 

The axe (Figs. 74: 7; 109: 4-7). Three axes were recovered in the caches at 
Agarshin in Xinjiang (blade fragments); in Tacheng, in Tuoli; in the east in 
Dzhimusaer, in China and Balikun. They belong to a specific Andronovo type. 
Axes are also found in southern Siberia (Fig. 109: 1-3). The main finds are 
concentrated in eastern Kazakhstan, Fergana, and Semirech'e (Fig. 74), as well 
as in the hoards from Shamshi, Sukuluk, Issyk-Kul', Alekseevka, and Tuksib 
(Chernikov 1960: 161; Kuz'mina 1966: 11-14, table II; 1994, fig. 43a; Kozhum- 
berdiev and Kuz'mina 1980, Avanesova 1978, 1991: 14, fig. 13). The axes of 
this type developed from Eastern European and early Andronovo lop-butted 
types. The type is characterized by the angular cross-section of the blade, a shaft- 
hole, with a terminal lug at the butt, and patterns in relief on the butt. In early 
samples comb patterns on the handle were not very prominent. The axe from Ji- 
musaer belongs to this type. In later samples the comb pattern is clearly marked, 
an ornament in the shape of a spruce or a net sometimes decorates the butt. Such 
ornament is present on the axes from Agarshin, Tacheng, Alekseevka, Sukuluk, 
Kirgizia and are kept in the University of Bishkek. There is a cross on an axe 
from the Altai and from Shemonaikhe (Fig. 109: 2, Frolov 1996: fig. 1.2). There 
is notching on the celts from Urlanova (Kiryushin and Ivanov 1996: fig. 2). 

The date of origin of the type is determine by hoards, the discovery of stone 
molds for axes in settlement # 16 in Kayrak-Kum, and two axes and ceramics 
with applied roller in the settlements of Bes-Tyube (Litvinsky 1962: 12, 213, 



RELATION OF THE ANDRONOVANS WITH XINJIANG AND CENTRAL ASIA 259 

table 36). Identical ceramics were found with an axe in the settlement of 
Krest'yanskoe IV in the Altai. From this region also come large sickle-scythes of 
Sosnova-Maza type in Timofeevskoe, and a sickle that has European parallels 
from Mayorovka (Ivanov and Isaev 1999: 83, fig. 1). This evidence places the 
date between the 13th and 9th centuries BC. Possibly, the type of the axe from 
the Baicaopo grave that belongs to the Western Zhou was created under the 
influence of Andronovo axes (Lin Yun 1986: fig. 55.3). This type is remarkable 
for having a straight butt, but just like the Andronovo ones, a cock-comb, an oval 
shaft-hole and a six-angled blade. The date of the Chinese sample does not 
contradict the chronology of the Andronovo types. 

The adze. In Xinjiang three adzes make up part of the Agarshin hoard, another 
three were found in Tacheng (Figs. 43a,b; 75: 18-23), and one to the east, in 
Urumchi. Flanged adzes were known in Siberia and northern Kazakhstan. The 
major discoveries are concentrated in the Altai, eastern Kazakhstan, especially in 
Semirech'e, and in hoards from Shamshi, Sukuluk, Sadovoe, Alekseevka and 
Tuyuk (Chernikov 1960: 164, table lxiv. 9, lxvii. 5; lxxvii. 3, 4; Kuz'mina 1966: 
18-20, table iii. 9, 10, 14-17; 1994). The date of this type is the 13th-9th century 
BC. This dating is based on the following archaeological evidence: hoards and 
the discoveries in Sary-Ozek, where adzes were found with sickles of the 
Sosnovaya-Maza type; finds in the settlements of Malokrasnoyarka, eastern 
Kazakhstan, Stepnyak, and others in northern Kazakhstan (Chernikov 1960: 82; 
Zdanovich S. 1979: 12) where adzes and ceramics decorated with applied-roller 
design were recovered. 

Socketed adze with bevel. A unique sample of this type was found in Xinjiang in 
Ksinuan. The tool without a doubt has a distant analogy to the single-cast 
flanged adzes and represents the local development of the type. A tool from 
Regar in Tadzhikistan is a distant analogue of the above-mentioned sample 
(Kuz'mina 1966: 23, table iv.7) 

Flat socketed chisel (Figs. 43a,b; 109: 12-16). Three chisels were part of the 
Agarshin hoard in Xinjiang. One was found in Tacheng; the other one was found 
in the east, in the Qizilchoqa settlement. The chisels have a round socket with a 
lug and a flat blade. The type is distributed across Eurasia from the northern part 
of the Black Sea region to southern Siberia. Examples have been found in 
numerous excavations in eastern Kazakhstan and in Semirech'e in the hoards 
from Shamshi, Sukuluk, Sadovoe, Alekseevka, and Tuyuk (Chernikov 1960: 70, 
80; Kuz'mina 1966: 26, table iii: 3-6). This category was developed in Eastern 
Europe at the beginning of the Early Bronze Age, and is represented by forged 
chisels with a twisted socket. The next step of development is represented by a 
chisel with a single -cast ornamented socket without a bevel such as the one from 
Rostovka (Chernykh and Kuz'minykh 1989: fig. 23. 5). The existence of this 
form led to the production of cast tools with a bevel and socket. The beginning 
date for the chisels was set by the discovery of a mold for chisels and a knife as 
well as by numerous Western analogies in the hoards of Sosnova-Maza, Krasny 
Mayak, Kardashinka and ceramics found in the settlements dating from the 
1 3 th-9th centuries BC. 



260 CHAPTER TWENTY 

A chopper-sickle (Figs. 43a; 75: 1-7). In Xinjiang two sickles were included in 
the Agarshin hoard; four were found in Tacheng; one in the east, in Hami; and 
one in the settlements at Qizilchoqa and Turfan. The sickles are massive, cast in 
an open mold, have a curved profile, a slightly curved inward blade and a wide 
heel with a hole sometimes decorated with a spruce design. Chopper sickles are 
distributed from the Volga, where they are represented in the Sosnova-Maza 
hoard, to southern Siberia. Many discoveries come from eastern Kazakhstan and 
Semirech'e including the hoards of Shamshi, Alekseevka, and Turksib 
(Chernikov 1969: 38, 44, table xxxvi.19; Kuz'mina 1966: 54, 55, table xi). The 
start date is the 1 3 th— 9th centuries BC which has been established on the basis of 
Western analogies and discoveries from settlements such as Malokrasnoyarka 
where ceramics with an applied-roller were found. The discovery of a sickle in 
the Kent settlement in central Kazakhstan (Fig. 42: 9) is of great importance, 
since it represents a type with well defined dates of origin: a notched spear, a 
bone cheek-piece, and a knob with a circular ornament (Varfolomeev 1991). 

The celt. Several types of celts have been found in Xinjiang. This tool category 
appeared in the Seyma complex and was further developed during the Saka 
period. In the late Bronze Age four-sided or oval celts with a beveled socket and 
no ornament replaced the celts of the Seyma type, six-sided, with or without eyes 
located below the socket and always generously decorated. 

Celts with oval sockets reinforced with a bevel and with two eyes are represented 
in the west, in Nileke; in central Xinjiang in the settlements of Xintala; and in the 
east, in Hami. The double -eyed type was formed in Seyma (Fig. 31) and is 
familiar from Andronovo archaeological sites, especially in eastern Kazakhstan 
(Kuz'mina 1966: 20-22; 1994: fig. 31; Chernykh and Kuz'minykh 1989: fig. 19, 
fig. 31). The Xinjiang samples represent a late degradation of the type; the 
proportions, the cross-section of the blade, and the geometric ornament are 
changed. The samples can be dated to the end of the final Bronze Age. 

The celt with an oval socket and a cast bevel originates from Kroran and is dated 
to the end of Bronze Age according to analogies with items in the Sadovoe hoard 
in Kirgizia (Kuz'mina 1966, table iv.13). 

Celt-hammers (Fig. 77: 13-15) were found in the Agarshin hoard and have 
analogies in the Kirgizian hoards from Shamshi and Sadovoe (Kuz'mina 1966: 
23, table iv.8). The late Bronze Age celts from Xinjiang are reminiscent of a 
stone mold fragment of a celt from Zhukaigou in northern China (Chzhun Suk 
Be 2000: fig. 13: 7). A celt with the characteristic Sadovoe net ornament on the 
eyed socket survived there. It possibly came to China from the west. According 
to the Andronovo analogies, the age of the site cannot be older than the 13th 
century BC. A celt with an open socket was found in Xinjiang, in Tacheng. It 
belongs to a type of tool dating from the final Bronze Age (Chernikov 1960: 84, 
table x. 3, 4; lxiv. 8) according to the discovery of Karasuk curved bladed knives 
in the Palatzy hoard. 



RELATION OF THE ANDRONOVANS WITH XINJIANG AND CENTRAL ASIA 26 1 

Celt-spades (Figs. 76, 77). Different tools of this category were excavated in 
Xinjiang: spade -shaped celts with a concealed socket (Type I) were found in 
Tacheng and Urumchi; in the settlement of Keremchi molds for celts with a 
protruding socket (Type II) were discovered. The centers of production of Type I 
that yielded tools with an internal socket were Fergana and Semirech'e. Single 
finds of such types were made in northern and eastern Kazakhstan as well as in 
southern Siberia (Chernikov 1960: 83, 84; table lxv. 4; Kuz'mina, 1966: 24-25, 
table v; Silvi Antonini and Bajpakov 1999, fig. 33). A mold of this type was 
found in the Samus' IV settlement (Chernykh and Kuz'minykh 1989: 154, fig. 
80: 7). The beginning date for the type is defined on the basis of the mold from 
Samus', a cemetery on the river Kurchum, with a knife with a hollow handle and 
curved blade of Karasuk type, and, most importantly, another example from a 
dwelling (#9) at the Chaglinka settlement where there were ceramics with 
applied-roller designs. 

Type II celt-spades with a protruding socket, sloping shoulders and blade 
facets are known from excavations at Lebedinovka, Dzhappa, in Fergana. A 
transitional type with extended facets but a concealed socket originated in 
eastern Kazakhstan and the Rostovka settlement in Siberia (Kuz'mina 1966: 22, 
table vi; Zadneprovsky 1996: 17, fig. 16; Chernykh and Kuz'minykh 1989: 63, 
fig. 22.4,5). The origin and chronology and their connections with the celt- 
spades of China have been discussed. A. V. Varenov (1999) stressed that the 
celt-spades of Semirech'e and the Altai have major differences from Chinese 
ones that have a six-faceted or square socket opposed to the round steppe model. 
Only the samples from the tomb of the Fu Hao grave at Anyang have a round 
socket, which points to a northern steppe origin for it as well as a number of 
other items, as K Linduff has suggested (1995). 

The origin of this type is not clear to me. If we refer to the fragment of the 
celt from the Abashevo settlement of Shigona and the Rostovka celt of this type, 
we can decide for a western Seyma origin, proposing that the tools of Type II 
were their derivatives and transitional from Type I. However, in my opinion, the 
possibility that celt-spades evolved from Chinese tools cannot be disregarded. 
The Chinese tools served a prestigious ceremonial function; generally, they were 
generously ornamented, sometimes having a blade inlaid with nephrite. 

The single-blade knife (Fig. Ill: 18-22). This category is widely distributed in 
Xinjiang, in the complexes dated to the end of the Bronze Age, and across 
Europe. The type of knife with a ring found in the Lanzhouwan settlement in 
Chaqimale, as has been mentioned before, dates back to the knives known from 
archaeological sites of the end of the 3rd-2nd millennium BC in northern China 
and in the Seyma-Turbino complexes of Eurasia. In the final Bronze Age these 
knives are found in settlements with ceramics with an applied-roller, particularly 
in the clearly dated settlement of Kent in central Kazakhstan (Fig. 42: 12; 
Varfolomeev 1991). 

The double-winged socketed arrow (Fig. 109: 13,14): Type I — with a protruding 
socket; Type II — with a concealed socket. Both types are widely known in 
Xinjiang in the settlements of Xintala, in Kersang, in the cemeteries of Nanwan, 
Qizilchoqa, Yanbulaq, Chaqimale, which are concentrated among agricultural 



262 CHAPTER TWENTY 

communities in the east. It is impossible, however, to say whether the objects 
belonged to the local population or to their enemies — the pastoralists. They have 
analogies with materials from late Andronovo sites. A number of arrows with a 
concealed socket have been found in Kazakhstan in the settlement of Kent, for 
instance, and in Fergana in the cemetery of Vuadil' (Kuz'mina 1966: 33-37, 
table 5) dated to the end of 2nd or the beginning of the 1st millennium BC. 

Double-winged arrows appear in the early Andronovo burial ground of Sintashta 
(Gening et al. 1992, fig. 171, 185, 186, pi. 45); they developed further in the late 
Bronze Age, when arrows with protruding as well as with concealed sockets 
become widespread (Avanesova 1975; 1991). The arrows in the Saka-Scythian 
period developed on the basis of Bronze Age types. It is important to realize that 
in Yanbulaq arrows of the Andronovo type are found alongside special Xinjiang 
types (Mei and Shell 1998: fig. 2.2-6). 

Socketed arrows (Fig. 54: 27-29) are not characteristic of China. While studying 
the military art of China, A. V. Varenov (1987: 199, fig. 7) connected their 
sporadic discovery in Anyang with those discovered in the archaeological sites 
of Baode and Shilou located in the northern periphery of the Shang Dynasty. The 
discoveries in Xinjiang contribute to this number. 

A single-bladed knife (Fig. Ill: 13) with a zoomorphic handle with the head of a 
deer from Chaqimale near Hami, as has been said earlier, belongs to a large 
group of knives present in the northern regions of China, in the Ordos, in Hebei, 
and Shanxi. Their appearance in rich graves in the Central Plain at Anyang, 
particularly in the Fu Hao tomb (Linduff 1997), and in the period of the Western 
Zhou are connected with northern peoples of the Karasuk culture who were 
ethnically different from both the Andronovans and the Chinese (Chlenova 1972; 
Wang Binghua 1996: 77; Lin Yun 1986: 245, fig. 49.17; Chzhun Suk-Be 2000, 
figs. 1.2, 2.1-5). As we said earlier the prototypes of these knives date to the 
Seyma-Turbino horizon. 

Analysis of the non-standard artifacts from Anyang permits us to conclude 
that they find their closest analogies in the complexes of eastern Kazakhstan and 
Semirech'e. The similarity in the composition of types in Semirech'e and the 
Agarshin hoard is used as a special proof of this statement — all the above 
mentioned tool types are present in the hoard of Shamshi along with other types. 

Chronological position of the hoards 

The chronology of the hoards is established in the following manner: 

1) Objects that have analogies in the western steppes are dated by the 
European chronology. These comprise chisels, sickles, and especially razors 
from the Shamshi hoard that had a very limited sphere of distribution. This 
evidence makes it possible to date these hoards back to the 13th -9th centuries 
BC (Kozhomberdiev and Kuz'mina 1980). 

2) The discoveries of the types present in the hoards in assemblages such as 
the following: an axe mold from Kayrak-Kum; axes in the Altai; adzes, chisels, 
sickles, single-bladed knives in the settlements of Fergana, and eastern, central, 
and northern Kazakhstan, where such objects are found with ceramics with 



RELATION OF THE ANDRONOVANS WITH XINJIANG AND CENTRAL ASIA 263 

applied-roller that are clearly dated to the 13th— 9th centuries BC on the basis of a 
wide set of analogies from the Danube and the Ukraine to Anatolia (Troy VIIB) 
and Iran (Tepe Giyan). This fact determines the original date of these hoards as 
well as the Agarshin hoard. 

The origin of the latter hoard, based on similarities with tools from Fergana 
and Semirech'e, can be defined as an import from Kirgizia, where all the 
categories of tools represented above have a long evolutionary development in 
the Andronovo culture. Ornaments from Andronovo complexes such as Fergana 
have analogies in Xinjiang. Beads were found in Xinjiang in the cemeteries of 
Tacheng, Nanwan and Qizilchoqa (the C 14 date is 1350-1000 BC). In the latter 
burial grounds, a chain of beads was sewn onto a pair of boots. Boots were a 
remarkable feature of Andronovo tribal outfits; they were made of leather or felt 
and were decorated with strings of beads (see chapter on 'Dress'). 

Mirrors (Fig. 54: 2, 16, 17): Type I — a mirror with a protruding handle was 
found in the burial ground of Qizilchoqa. This type is present in Kirgizia in the 
hoards of Shamshi, Sukuluk, and Sadovoe (Fig. 110); in Fergana in the settle- 
ments of the Chust culture (Zadneprovsky 1962: 68, table xx. 4,5; Kuz'mina 
1966: 68, table xiii). Type II — a mirror with a loop handle was found in the 
Yanbulaq cemetery (Mei and Shell 1998: fig. 4: 1). Analogies to this type have 
been found in burials of the late Andronovo culture in Smolino (Urals); in 
Elovka (Siberia); in Kara-Kuduk (southern Kazakhstan); in Yapagi (Fergana); in 
Kul'say, Kizylbulak (Semirech'e); in Muminabad (Uzbekistan), and other pla- 
ces. Examples are also present in the burials of farmers at Bustan. A mold was 
found in Fergana in the settlement of Dal'verzin that belongs to the Chust culture 
(Gorbunova 1995; Mar'yashev, Goryachev 1999: fig. 5.5; 9.15; Zadneprovsky 
1962: table xx: 5). 

The round, flat or concave mirrors appeared in the farming cultures of 
southern Turkmenia, Bactria and Margiana in the 3rd millennium BC. This type 
diffused further to the Zaman-Baba culture in Central Asia, and later to the 
Sintashta groups in the Urals. During the Andronovo period this type was known 
only in Central Asia (Muminabad). Different types of mirrors appear in the late 
Bronze Age: round with a protruding handle, square with a rounded loop. They 
are present in Kazakhstan together with ceramics with an applied-roller and in 
hoards from Semirech'e. Two mirrors with a loop handle were found in Shamshi 
(Fig. 33; 43a: 5). 

In China this type can be found in the north, in Inner Mongolia and some- 
times in Shang burials; four examples were found in the tomb of Fu Hao, a 
consort of Wu Ding (13th BC) at Anyang (Linduff 1994, 1996, Lin Yun 1986, 
fig. 51.5, 8, 9) where other items of northern origin are also present. A mirror of 
a smaller diameter was discovered at Houjiazhuang in grave 1005, dated to 
1300-1028 BC (Juliano 1985: 38-43, fig. 1.4). These samples are "decorated 
with geometric ornament, foreign to the Shang art style, but with analogies in the 
decoration of Andronovo artifacts" (Kuz'mina 1988: fig. 2). 

Round plates decorated with punched ornament along the edge from 
Yanbulaq (Mei and Shell 1998: figs. 5, 14) are also characteristic of Andronovo 
complexes. This fact gave me reason to accept the opinion of Lin Yun (1986), K. 
Linduff (1995; 1997; 2000a) and A. Juliano ( 1985) who support a northern and 



264 CHAPTER TWENTY 

western origin for the mirrors in China. Since the mirrors are concentrated in 
archaeological sites of late Bronze Age in eastern Turkestan and Semirech'e it 
would be possible to imagine that they reached China from that particular region. 
However, the discoveries of flat mirrors with a geometrical ornament in Erlitou 
and Yamatay in Qinghai are dated to a very early period. This makes the 
Andronovo hypothesis quite plausible. L. Fitzgerald-Huber (1995: 53, fig. 10a) 
suggests that they originated in Bactria. 

Earrings (Figs. 54; 1 10). The type made of a spiral cone of wire is present in the 
burial ground of Yanbulaqe (Mei and Shell 1998: fig. 6.2). Identical earrings are 
found in the cemeteries of Ketmen' -Tyube and Tamgaly in Semirech'e and 
Arsif; and at Yapagi and Kashkarcha in Fergana. The latter samples are made of 
gold and silver (Kozhemeko 1960, Gorbunova 1995; Rogozhinsky 1999, fig. 17: 
4) and are present in the Ob region (Avanesova 1991: 57-58) and central 
Kazakhstan in the cemeteries of Sangru II and Shoindykol (Evdokimov and 
Usmanova 1990: 78, table II) 

Earrings with trumpet ends often made of gold or gilded constitute a specific 
Andronovo type. They are widely distributed in Andronovo areas and are 
connected mainly to the archaeological sites of the Fedorovo type. They are 
found in the cemetery of Borovoe in northern Kazakhstan, Sanguyr in central 
Kazakhstan, Tautary in southern Kazakhstan, Maly Koytas, Kytmanovo, and 
others in Siberia (Avanesova 1972; 1991: 43, 53). Many discoveries of this type 
of ornament have been made in eastern Kazakhstan: at Kyzyltas, Predgornoe, 
Zevakino, Berezovskiy (Tkachev and Tkacheva 1996: figs. 1. 3, 2. 4); and in 
Semirech'e: Mynchunkur, Tamgaly IV, VI, Kul'say, Uzunbulak, Kyzylbulak; 
also in Tash-Tyube II, Tegirmensay (Kuz'mina 1966; 75, table xiv; Rogozhinsky 
1999: fig. 5: 1-3.9) and in Fergana: Vuadil', Dakhana (Gorbunova 1995). In 
Central Asia earrings with a trumpet end are found in the cemeteries of 
Tandyryul and Dashti-Kozi in Tadzhikistan, Muminabad, Bustan VI (Fig. 70: 9), 
in the peasant settlement of Sazagan (Kuz'mina and Vinogradova 1996, fig. 6; 
Avanesova 1997, fig. 13.18). In the West, I have information about only one 
golden earring from my excavations in the Baytu cemetery. N. A. Avanesova 
dated this type back to 14th 13th century BC. A. A. Tkachev and N. A. 
Tkacheva suggested that this type is older based on an association with a knife 
that belonged to the Seyma period. The dating of the Baytu complex reinforces 
the possibility of dating the beginning of earring production. 

Most likely the chronological boundaries of this type are the 14th to 1 lth- 
10th centuries BC. The first date is defined by the associated discoveries of 
earrings and a socketed arrow in Vuadil' and bracelets in Sanguyr II. Earrings 
with trumpet ends are known in China: in the northern zone, in Dongbei 
(Larichev 1959), in Liulihe, Xiaoguanzhuang (Lin Yun 1986: 248, fig. 50.8, 9; 
Bunker 1993: 37, fig. 4; 1998: 611, figs. 3, 4, 1 1). E. Bunker has shown that the 
elite of the Shang in Anyang did not use gold, because nephrite was the 
prestigious material. This fact points to the western origin of golden earrings 
with funnel ends found in China. They could reach the elite of the Northern 
peoples from the Altai or, which is more likely, from Semirech'e and Fergana. 
This logical hypothesis contradicts the fact that earrings with trumpet ends from 



RELATION OF THE ANDRONOVANS WITH XINJIANG AND CENTRAL ASIA 265 

Xiajiadian (Linduff 1996a: fig. 4: 2, 4) are dated back to 2000-1600 BC, and are 
thus older than the Andronovo ones. 

The analysis of Semirech'e bronzes shows that this region was part of the 
Andronovo zone of influence. The number of Andronovo types decreases to the 
east. The influence of the eastern Kazakhstan and especially Semirech'e centers 
was crucial for the development of metal working in eastern Turkestan. The 
proof of this statement is the similarity of tools, casting technology (use of 
bipartite stone molds), and the metal composition of lead bronze. 

The distribution of Andronovo types in Xinjiang was the result of the 
following: 

1) Down-the-line exchange. 

2) The arrival of bronze smiths from Semirech'e, probably proven by the 
Agarshin hoard with analogies in the Shamshi hoard. 

3) Migrations to the east of the Andronovo population. This is supported by 
the cemeteries of Sazi and Tacheng. 

Having received an impulse from Andronovo and maintaining contacts with 
the west, the population of eastern Turkistan put into operation their own metal 
workshops, which can be seen in the discovery of a copper ingot in Tacheng and 
molds in Keremchi and Xintala. The development of local types such as socketed 
adzes with a bevel and double-eyed celts underscore the idea of a local industry 
as well. The end of the 2nd millennium BC saw the beginning of separation of 
metal workers from the rest of the community in the steppe. They were making 
standard types of tools for exchange, as can be seen from the hoard of Sukuluk II 
with its 17 sickles. Other hoards found in Semirech'e and Agarshin have a 
similar character. The fact that they were hidden underground reflects the hostile 
environment of the steppe at that time. 

This change can possibly be connected to important ecological and historical 
events in Asia at the end of 2nd millennium BC. The rapid drop in temperature 
impelled part of the population to adopt a new mobile pastoralist life-style. Cattle 
were an easy target and military action and social differentiation were initiated as 
a result. In times of war, valuable items were buried. The mobile life-style 
encouraged cultivation of new ecological zones and reinforced cultural and 
ethnic contacts. 

This analysis shows that the part of the Silk Road (Map 16) leading from 
Xinjiang to the steppes of Eurasia was mastered already by pastoral groups in the 
Bronze Age. The peoples of the Andronovo culture initiated the process. Major 
objects of export were ready-made metal tools. These contacts were especially 
active in the Bronze Age, in the 13th and 9th centuries BC. 



The classification of Andronovo material reveals complex historic processes 
of autochtonous development, migration and integration. It is probable that a 
migration of the Alakul' people in the second quarter of the 2nd millennium BC, 
from western into central Kazakhstan caused part of the Fedorovo migration to 
eastern Kazakhstan and south Siberia. Another part of the Fedorovo people was 
assimilated by the Alakul' people, and as a result of this integration, syncretic 
complexes were formed. Moreover, the Sol'-Iletsk and Kozhumberdy types 
formed very early: in central Kazakhstan long contacts were of different 



266 CHAPTER TWENTY 

character, which conditioned the mixed character of the Atasu type; the 
Semirech'e type was probably formed rather late, as a result of population 
migration from central Kazakhstan. 

A review of the numerous sites of these syncretic types, united in an integral 
chain across the major part of Andronovo territory permits one, following A. A. 
Formozov, to speak about the existence of an Andronovo culture unity. 

The Andronovo materials surveyed so far provide a database for the analysis 
of the material cultural of the Andronovo tribes. Of central importance in the 
development of the stockbreeding cultures of the steppe ECT was the transition 
to a nomadic type of economy, the most basic innovation of Old World culture. 
That is why the study of the ECT of this region is of primary interest for 
examining its paleo-economic development and revealing the processes 
involving the emergence of mountain pastures and then the creation of nomadic 
stockbreeding in Eurasia, which explains the dynamics and intensive 
assimilation of new territories. A. J. Toynbee, who dedicated his A Study of 
History to the philosophy of history and who greatly influenced the development 
of modern historical schools, expressed the opinion that steppe peoples were a 
catalyst of all the processes in the history of civilization. Similarly, the head of 
the French 'Annales' school, F. Braudel (1969) considered the Eurasian steppe 
as a flashpoint that saw the explosion of steppe pastoralists cultural from 
Germany to China who punctuated the slow process of evolution in the Old 
World. 

Finally, the separate categories of the material culture of the Andronovo 
tribes and their comparison with language data and historical traditions of the 
ancient Indo-Iranians will help resolve the problem of locating their homeland, 
reject speculative constructions, and give the floor to the Aryans themselves. 



CHAPTER TWENTY -ONE 

CULTURES OF NORTHERN BACTRIA 
IN THE LATE BRONZE AGE 

BMAC sites in Uzbekistan 

As already mentioned, the Sapalli variant of the BMAC formed no later than the 
first quarter of the 2nd millennium BC in northern Bactria as a result of 
migration from the south-west (Map 17). It comprises the following sites: on the 
Sherabad river — the cemeteries of Bustan and Dzharkutan and the settlement of 
Dzharkutan (the upper horizon; Fig. 88), on the Surkhan-Darya river — the 
settlement and cemetery of Mollali (Pugachenkova 1972; Belyaeva and Kha- 
kimov 1973; Askarov 1977, 1989; Askarov, Abdullaev 1983; Askarov and Io- 
nesov 1991; Rakhmanov 1982; Rakhmanov and Shaydullaev 1985; Avanesova 
1992; 1995a, 1995c; 1996; 1997; 2002; Shirinov and Baratov 1997). A. Askarov 
assigns the monuments to the Sapalli culture, the other researchers attribute them 
to the Namazga VI culture or the Bactria-Margiana Archaeological Complex. 
Recently A. Askarov and T. Shirinov (1993: 42) suggested that a Namazga 
historical and cultural community should be distinguished comprising the entire 
south of Central Asia, north-eastern Iran and northern Afghanistan. As men- 
tioned above, I deem it more appropriate to use the term 'the Bactria-Margiana 
Archaeological Culture (BMAC)', considering the similarity of the historical 
destinies of the population of this region and the considerable differences in the 
cultural development from western Turkmenistan and eastern Iran. 

The relative chronology of the sites was developed by A. Askarov (1977: 60- 
63), who singled out the following stages: I - Sapalli, II - Dzharkutan, III - 
Molalli. Later the Molalli stage was divided into III - Kuzali and IV - Molalli 
(Askarov and Abdullaev 1983: 40) and further into IV - Molalli and V - Bustan 
(Rakhmanov 1987: 13, Shirinov and Baratov 1997: 88). The accepted chron- 
ology of the stages are as follows: Kuzali: 1300-1200 BC, Molalli: 1200-1050 
BC, Bustan: 1050-950 BC. They were succeeded by the culture of painted 
ceramics Kuchuk II-Yaz I. 

The relative chronology is based on the stratigraphy and dynamics of the cer- 
amic complexes. The absolute chronology is contestable (see below and Askarov 
and Shirinov 1993: 84-92 for a discussion). One can trace the genetic succession 
of the culture's stages, but at each stage some innovations come into being. 

In the Sapalli period burials were placed within the precincts of the 
settlement in a ground pit or a catacomb blocked up with brick; burial was in the 
flexed position, men on their right side, women on the left. Occasionally, the 
deceased person was placed in a large vessel (Khum). Cenotaphs are familiar 
containing a clay dummy or a ram' s skeleton instead of the deceased (Askarov 
1973). 



268 CHAPTER TWENTY-ONE 

Beginning with the Kuzali stage relations with the Andronovo tribes are 
recorded, reaching their apex in the Molalli and Bustan periods (Kuz'mina 1972; 
Askarov 1989; Rakhmanov et al. 1985; Avanesova 1992; 1995; 2002). The 
burial ritual has changed. Discrete cemeteries come into existence. Burials are 
found in pits or catacombs, blocked by bricks or stones. The dead lie flexed, men 
on their right side, women on the left, head oriented predominantly to the west. 
There are also dismembered burials. The graves contain wheel-made BMAC 
vessels, sometimes clay or metal votive articles (vessels, spoons), as well as 
ovicaprid bones. 

The fire -cult becomes widespread. Cremation has been revealed in the 
cemeteries of Dzharkutan IV, Bustan 3 and 6. In the Bustan 6 cemetery 17 of the 
130 burials are interred according to the cremation rite. It is notable that it is in 
these burials that Andronovo ceramics of the Fedorovo type were found (Fig. 70; 
Avanesova 1995, fig. 12). 

Evidence of a fire -cult has been discovered in 82 burials: fires, hearths, pits 
and vessels with ashes, ashes in the entrance to the catacomb with inhumation, 
small round clay altars (one — with a hand-made bird's head) with coals and 
ashes. In some constructions there were found pieces of red ocher and white 
plaster. Cenotaphs, occasionally containing a sheep skeleton, are numerous. In 
17 graves 22 unbaked clay anthropomorphic figures were found. 

Of considerable interest is the discovery in Bustan of three fires and in their 
vicinity three cists — one rectangular and two trapeziform. They are formed from 
bricks with the help of a clay solution and show signs of repeated fires that were 
then put out by water and oil. 

Fire was made three times in the ritual and the burnt layers were covered over 
with sand and white gravel. In the boxes there were found calcined bones of a 
man and ovicaprid, the leg of a Molalli type wheel-made vase, a fragment of an 
Andronovo vessel, a cylindrical urn with a conical lid and a bronze earring with 
a hook. N. Avanesova (1995: 33-35, fig. 3, 4) interprets these constructions as a 
crematorium. 

The Andronovo provenance of the fire -cult and the cremation rite is beyond 
dispute. Bustan's type of fires and crematoria goes back to the cult constructions 
discovered in the Fedorovo-type cemetery of Kinzerskiy in the southern Urals 
(Kuz'mina 1975: 222, 227). In the center of the cemetery in mound 33, the lar- 
gest one, in a pit under the apex, there was a fire and animal sacrifices; another 
fire and sacrificial pits were revealed inside the circular rampart. In burial mound 
17 under the bank surrounded by a circle of stones placed in horizontal rows, 
there was a ring of red clay and ocher; and deeper yet, there was a solid red 
circle with a pit in the center. The pit's walls were burnt as a result of repeated 
strong fires. At the bottom of the pit there were found fragments of two vessels 
(one bearing a swastika on the base) and an animal's shoulder-blade; above there 
was a layer of pure ashes covered over with red clay and yellow ocher with 
traces of water and oil poured on. I interpret this construction as a crematorium. 

In central Kazakhstan in the Shapat cemetery of the Atasu late stage type 
alongside the inhumation rite there was discovered a construction related to the 
fire -cult. It was a ground pit (2.0x1. 6x0. 5m) with a ledge in the western section 
0.5m in height and 0.2m in width. It was filled with bones of cattle and 



CULTURES OF NORTHERN BACTRIA IN THE LATE BRONZE AGE 269 

fragments of ceramics. The pit's bottom was burnt due to repeated fires and 
covered with a thick layer of coals and ashes (Varfolomeev 2002, part 2: 107). 

In central Kazakhstan there are numerous occurrences of altars in the form of 
stone rings and fences containing a hearth with layers of charcoal and traces of 
water pouring, potsherds and wooden tools. Sometimes they occur near menhirs 
crowned with rams' heads (Margulan et ah 1966: 154-156; Margulan 2002: 331- 
335). In the cemetery of Bustan there were also discovered leg bones of a horse 
and the clay figure of a horse (Avanesova 1995, fig. 10), which testifies to the 
spread of the cult of this animal to Bactria. It is alien to the world of the farmers 
and is particularly characteristic of the Andronovans (Kuz'mina 1977). 

The Andronovo ceramics are represented in Bustan (Fig. 70) by: 1) distin- 
ctive pots of the Fedorovo-type with a high neck and a rounded shoulder, ornam- 
ented in plain or indented stamp in oblique triangles over the neck and zigzag or 
x-shaped figures over the shoulder; 2) jars with convex sides and an unaccen- 
tuated narrow neck, decorated in zigzags underneath the rim, and 3) a vessel with 
a convex body, a short rim and ornamented in triangles over the shoulder 
(Avanesova 1995, fig. 12) (the last shape is typical of the Late Bronze Age). Of 
interest is also a vessel imitating the Fedorovo-type fluted ornament (Avanesova 
1997, fig. 14: 1). 

In the grave there were found distinctive Andronovo ornaments (Fig. 70: 
1,2,5,6,8,9): bracelets with open ends, made from a convex plate, bracelets with 
conical spirals, trumpet -shaped earrings, plaques with punched ornamentation 
and plaques with a loop, various beads as well as a nail-like and a bi-spiral 
pendants, a pin and lazuli beads (Avanesova 1997, fig. 3: 4; 10: 1-4; 13). 

No less conspicuous are the traces of contacts with Andronovans in the 
Dzharkutan complexes of the Late Bronze Age (Fig. 70: 28-30; Askarov 1989). 
In the cemetery of Dzharkutan IV cremations have been revealed, and in a grave 
of the Kuzali stage — an Andronovo bracelet. At the Dzharkutan settlement two 
large stone cists were investigated, 4x0. 7m and 0.5m deep, formed from vertic- 
ally dug-in stone slabs and covered over with stone slabs. The cists contained 
calcined human bones, charcoal, ashes and small fragments of Molalli type 
ceramics. 

At the settlement of Dzharkutan A. Askarov discovered a large temple 
complex including storage and cult facilities. The area of the settlement is 35 x 
35 m. It is a 25m long corridor, a small inner yard with a sacred well, a sacral 
platform on which the fire -altar and the depository of the holy ashes were 
situated. In its center there is a rectangular reservoir with a 10cm layer of ashes 
whose bottom was covered with a layer of fired red clay containing charcoal, 
bones, and potsherds. On the top of it was a layer with animal bones. In the 
upper horizons of the reservoir fill wheel-made ceramics of the Molalli type 
were found and sherds of Andronovo pottery (Fig. 88). Of particular interest is 
the discovery of mobile ceramic altars and cult vessels, made on the potter's 
wheel, light-slipped, but decorated in plain stamp and incision — ornaments 
typical of the steppe Bronze Age cultures. Similar pottery, apart from the temple, 
was found also in other sections in the overlying layer of the settlement (Rakh- 
manov 1982: 45-119; Rakhmanov and Shaydullaev 1985; Askarov 1989: 165; 
Askarov and Shirinov 1993: 61, 66-70, 82; fig. 22 I: 4-6; 43 I, III; 44 I: 17-19; 
46-1; 64 I: 3; 74). There is a single fragment with ornamentation in isosceles 



270 CHAPTER TWENTY-ONE 

triangles, executed with an indented stamp (Askarov and Shirinov 1993, fig. 64 
I: 3). The ornamented vessels are represented by supports with images of a swas- 
tika and an eight-spoked wheel (Askarov et al. 1993, fig. 43 I: 3; II: 10, 12; II: 
4,5), pots with a wide neck and out -turned rim, underneath which there is a zig- 
zag or pendant triangles hatched horizontally in indented stamp (Askarov et al. 
1993, fig. 44: 18, 19). There are vessels with an out-turned and thickened rim 
which bears oblique notches, sometimes overlapping to form rhombuses or a net; 
often there is also a zigzag underneath the ridge (Askarov et al. 1993, fig. 22: 4- 
6, 461: 5-7; 12-14; Rakhmanov 1982, fig. 3, 5, 8). These decorations imitate the 
ornament of the steppe ceramics of the Late Bronze Age. Of the most profound 
interest are the wheel-made pots and particularly the cylindrical vessels with a 
thickened rim and an applied-roller in the upper section; the roller and the rim 
are ornamented in oblique notches or a net (Askarov et al. 1993, fig. 46 I: 1-4; 9- 
11; 74; Rakhmanov 1082: 16, fig. 1, 6). The applied-rollers with notches are a 
distinctive feature of the Alekseevka-type Andronovo ceramics. Their appear- 
ance on the pottery of Central Asian farmers reflects the interaction between the 
steppe and farming population in northern Bactria. To the late Andronovo forms 
also belong the single -bladed knives from Dzharkutan (Askarov et al. 1993, fig. 
62 I: 2; 67 IV), and the dagger from Vakhshuvar (Fig. 53: 1; Rtveladze 1981). 

What were the relations between the migrants — the Andronovo population 
and the indigenous farmers? A. Askarov (1989: 164) held that those buried 
according to the cremation rite were 'male and female concubines', later he came 
to the conclusion that the aborigines "would not adopt the backward material and 
technical basis of the northerners, syncretization occurred in the ideological con- 
cepts of these tribes." N. Avanesova (1995: 37) believes that the Andronovans, 
who brought with them the cremation rite, formed in the BMAC community a 
privileged social group. There is every reason to endorse these conclusions. 

Northern Bactria provides a unique opportunity to trace the southward migra- 
tional process of the Andronovo population and its assimilation with the locals. 
Since the material culture of the aborigines was highly developed and adapted to 
the ecological environment, the newcomers adopted in its entirety the complex 
of their material culture, while retaining their ethnical distinction in the most im- 
portant sphere — ideology: in the cults and burial rite. As is well known, the prin- 
ciple condition for maintaining ideology in traditional culture is the preservation 
of the language which conveys mythological concepts and ritual texts. For the 
understanding of the ethno-cultural processes in northern Bactria in the Late 
Bronze Age of crucial importance are the materials of the Dzharkutan settlement. 
They indicate that the northern migrants — Andronovans — not only preserved 
their language and beliefs but also disseminated their ideology among the aborig- 
inal population. This is the only way to explain the appearance in the sanctum 
sanctorum — the temple, the altars and cult vessels, the Andronovo cult ornam- 
ents. There are no military clashes or traces of fires in northern Bactria. Ob- 
viously, the relations between the two groups here, as distinct from Turkmenia, 
were peaceful. Numerous ethnographic and historical facts evidence that in the 
process of peaceful assimilation of two ethnic groups there was at first a period 
of bilingualism, then one of the languages would absorb the other. (This was the 
pattern, for instance, of the turkization of the Iranian-speaking farmers in Central 
Asia). Since in the assimilation process in northern Bactria it was the ideological 



CULTURES OF NORTHERN BACTRIA IN THE LATE BRONZE AGE 27 1 

concepts of the Andronovans that took the upper hand, it means that their 
language conveying ideology and ritual activity became the winner too. 

If the hypothesis of the Indo-Iranian attribution of the Andronovo culture 
advocated in this work should be accepted, then the materials of southern 
Uzbekistan demonstrate the major pattern of the distribution of the Indo-Iranian 
language in Bactria. 

Along with this type of migration (apparently, the principal one), there 
existed other models of ethno-cultural interaction. These models are prominently 
represented in the materials of southern Tadzhikistan. 

Northern Bactria and Tadzhikistan 

The natural conditions of Tadzhikistan are very peculiar: the territory is cut 
through by the tall ridges of the Pamir mountains and wide river valleys in the 
south. In ancient times irrigation farming was impossible in the valleys: it came 
into existence only as technology developed in the Early Iron Age. Ecologists 
have established that in the Late Bronze and Early Iron Age the climate was 
milder and more humid and the area of the fescue steppes vaster than today with 
wormwood-and-grass meadows being predominant (Litvinsky 1972: 181; Spiri- 
donova 1989: 16). In the Bronze Age the valleys with their grassy meadows and 
the warm climate constituted ideal pastureland that enabled one to keep cattle in 
winter. In summer they could be driven up to the high mountains. This deter- 
mined the primary orientation of the regional economy and the development of 
transhumance. 

At the same time, the bogar (dry, semi -irrigation) type of farming was quite 
feasible on the mountain terraces with their brooks and abundant precipitation 
(P'yankova 1998: 163). Annual precipitation in the south is just 140mm, whereas 
in the mountainous areas it reaches as much as 800mm. 

As a result of the diversity of natural and climatic conditions there arose a 
diversity of economic and cultural types that emerged in the process of 
adaptation to the ecological specifics by the representatives of the various 
archaeological cultures that populated Tadzhikistan. 

B MAC Sites in Tadzhikistan 

During its late stages BMAC populations pushed to the east and settled in the 
piedmont valleys of Tadzhikistan. The culture is represented in the upper reaches 
of the Kafirnigan river: by the cemeteries of Zarkamar and Tandyryul, burials in 
the late cemetery of Tup-Khona (D'yakonov 1950: 167, 176) and Kara-Pichok, 
and in the south of the valley — by two burials in Shah (Litvinsky and Sedov 
1983: 80); in the upper reaches of the Vakhsh and Tairsu rivers — by the settle- 
ments of Dakhana, Teguzak, Kangurt-Tut; on the Kyzylsu river — by the cemete- 
ries of Parkhar and Khodzha-Goib (P'yankova 1994: 355-372; 1998: 163-170; 
Vinogradova 1994: 29-47; 1997: 41; 2000: 89-109; 2001: 199-201; Gotzelt et al. 
1998: 115-144; Lombardo 2001: 271-280). The north-eastern outpost of the 
BMAC is Dzham (future Sogd region), not far from Samarkand, where BMAC 
vessels have been recovered, mirrors (a flat one and a round one with a geomet- 
rical ornament), a pin with a deer's figure, a double-sided sleeved adze as well as 
Andronovo type articles: a sleeved double -bladed arrow, bracelets, beads and 



272 CHAPTER TWENTY-ONE 

trumpet-shaped earrings (Fig. 84: 1-11; Avanesova 2001). To the same culture 
also belongs the layer of stage B at the settlement of Shortughai in Afghanistan 
(Francfort 1989). The stratigraphical data of Shortughai indicate that the BMAC 
period, containing the Fedorovo-type Andronovo ceramics (Fig. 91: B), is chro- 
nologically subsequent to stage A of the Harappa culture and is cut through by 
Bishkent-type burials, synchronous with the Fedorovo-type Andronovo ceramics 
and the late Andronovo ceramics with an applied-roller. At the settlement of 
Teguzak the lower layer belongs to the Neolithic Hissar culture. It is overlaid 
with a principal layer of BMAC which contains Andronovo ceramics. Above it 
pottery of the Yaz I - Kuchuk II type was found (Vinogradova 2000: 93, 94). 
The most comprehensive information has been recovered from the settlement of 
Kangurt-Tut: the Hissar layer is overlaid by the BMAC principal layer 
containing materials of the Andronovo (Figs. 89; 90) and Bishkent cultures, and 
on the top of it Yaz I - Kuchuk II ceramics were found. At the Neolithic sites of 
Tutkaul and Sayed, in the upper part of the Hissar culture layer, Andronovo 
ceramics were discovered. 

In Tadzhikistan small settlements are situated on mountain slopes. The hou- 
ses are located on the terraces: type I — a house measuring 18 x 6.7 m (Kangurt- 
Tut); type II — smaller constructions. The foundations of the walls are formed 
from horizontal rows of stones with the help of a clay mortar; above them there 
likely went a wall of clay with straw (pakhsa). The floor is occasionally faced 
with pebbles; round hearths were discovered. Type III houses were proto-yurts 
faced with stone. The large houses with stone foundations and proto-yurts are 
typical of late Andronovo architecture (Margulan 1979: 299). 

The economy was complex. The bogar type of irrigation is documented by 
grain imprints of different varieties of barley and wheat, querns and grain- 
storage pits. Cattle (52%) and ovicaprids (19.5%) were reared as well as horses 
(19%), donkeys (5.6%) and camels. 

The stone articles maintain the traditions of the Neolithic Hissar culture. The 
development of metal-working is borne out by the findings of stone casting 
molds. In Teguzak there were two molds for knives and awls, in Kangurt-Tut — a 
mold for a dagger with a support as well as a sickle with apertures in the handle, 
two single-bladed tailed knives with a bent back blade, ribbed and shafted dag- 
gers (Fig. 53: 25, 31). The articles are made of tin bronze, on occasion, with an 
admixture of lead. Apparently, the imported raw material came from the Zerav- 
shan where tin was mined in Zorabulak (Askarov and Ruzanov 1990). Typolo- 
gically and in terms of the alloy composition, the majority of the articles belong 
to the Andronovo metallurgical province (Kuz'mina and Vinogradova 1996). 
Ceramic production was highly developed. The vessels are made on the potter's 
wheel of clay with an admixture of fire-clay, baked in two-tiered kilns and 
coated with white, seldom red, slip and sometimes with vertical and horizontal 
polishing. Several types of ceramics are distinguishable: a vase on a foot, a large 
deep basin, a vessel with a globular body, a narrow neck and occasionally a 
truncated conical base area separated by a rib ('with a kneed bottom'); a pot with 
convex sides and a bent back ridge of small diameter, miniature pots, and also 
hand-made gray kitchen pottery and vessels for storage (Khum). This complex 
belongs to the Molalli and Bustan stages of the Sapalli variant of the BMAC. 



CULTURES OF NORTHERN BACTRIA IN THE LATE BRONZE AGE 273 

The Andronovo pots are hand-made of clay with a heavy admixture of 
quartz, feldspar and limestone; they are fired gray or brown (Figs. 48; 89). The 
shoulder is rounded, the ridge is equal to or, more often, smaller in diameter than 
the neck. The majority of vessels are devoid of decoration. Ornament is applied 
in plain or large indented stamps over two zones: on the neck and shoulder. The 
ornament comprises horizontal zigzags, oblique triangles, including elongated 
and sometimes opposite ones, and pennants. Occasionally one finds ornamenta- 
tion with dimples, imprints, and flutes. At the settlement of Dakhana the Molalli 
stage wheel-made pottery accounts for 86%, hand-made comprises 14%. Andro- 
novo-type pottery is not present. In Kangurt-Tut the Molalli and Bustan wheel- 
made pottery claims 53%, hand-made — 47% with the latter being fractionally 
represented by ceramic fragments of the Andronovo and Bishkent cultures. At 
the settlement of Teguzak in one of the prospecting shafts gray hand-made 
ceramics reaches as much as 43% (L.T. P'yankova assigns it to the Andronovo 
culture), but only a small part of it is ornamented. 

These facts point to the increasing interaction between the farming and 
pastoral population. 

The population's ideological concepts are reflected in clay anthropomorphic 
figures and a horse's head (Fig. 90: 15-17). Of particular interest are the ritual 
complexes. At Teguzak a ritual fire was cleared, fenced off by a stone wall. A 
large (7x3. 6m) concentration of ash contained broken vessels; the rectangular 
altar towering in its center consisted of layers of orange -baked earth; on the 
outside it was coated over with clay. This complex resembles the fires of the 
cemetery from Bustan 6. 

At the settlement of Kangurt-Tut four ceramic kilns were discovered. Two of 
them are circular, about lm in diameter and 0.7m deep; they were used to bury a 
human skull and leg bones. In kiln #4 the skull belongs to a male and there were 
animal bones found. In the chamber of kiln #3 the leg of a vase was discovered, 
the remains of a funeral feast over the burial revealed cattle and ovicaprid bones. 
As determined by L. Yablonsky, the female skull is close to the Andronovo 
skulls of eastern Kazakhstan and those of the Tazabagyab culture (I express my 
gratitude to N. Vinogradova for this information). Nearby in a pit was the partial 
burial of a male of huge build lying flexed on the left side, head to the west (ribs 
and vertebra are absent, the leg bones have been moved). It appears it also 
belongs to an Andronovan. 

The ritual of the BMAC cemeteries in Tadzhikistan is typical of the Molalli 
stage. There are dismemberments and children's burials within the precincts of 
the settlements. At the settlements of Kangurt-Tut and Tandyryul catacombs 
were revealed blocked up with stones rising above the ground or with clay. 
There have also been discovered ground pits and pits with an entrance shaft. On 
the floors of the chambers there are traces of pools, probably, resulting from 
pouring ritual water; also found are pieces of ocher and clay as well as ovicaprid 
bones and foot-bones. Cenotaphs are numerous, accounting in Kangurt-Tut for 
90% of the graves. Some of them contain an anthropomorphic figure of unbaked 
clay, and in one grave there are figures of a man, woman and child between them 
(Fig. 90). There are paired burials (Shah, Nurek). The dead lie flexed, men on 
their left, women on the right side, head most often oriented to the north-east or 
south-west. Some dismemberments are also recorded of which only a skull and 



274 CHAPTER TWENTY-ONE 

the long bones of the arms and legs are preserved. In addition, there were found 
burnt bones and charcoal. 

Of special interest is the discovery of grave 25 in the Tandyryul cemetery 
next to the BMAC burials. In a catacomb blocked up with stones there was found 
a woman's burial accompanied by ocher, a bead bracelet, a trumpet -shaped 
earring, an oval pendant, a lazuli bead, ribbed paste beads, and a hand-made gray 
pot with an everted rim and a rounded shoulder ornamented with flutes. Another 
hand-made pot with flutes was found in grave 2 together with nine wheel-made 
BMAC vessels (Kuz'mina and Vinogradova 1996: 39, 41, fig. 4: 2, 12). A frag- 
ment of an Andronovo vessel is present also in one more grave. Vessels with 
flutes are characteristic of the Fedorovo sites over the whole Andronovo range 
(Figs. 19: 11, 15; 20a: 3; 21: 13; 26: 3), while trumpet-shaped earrings are 
peculiar to its eastern territory (Kuz'mina 1966: 75, 76; Avanesova 1975: 67). In 
the west was found the only gold specimen in the Kupukhta cemetery (Kuz'mina 
1963b). This type of ornament is familiar in sites of the Andronovo and BMAC 
circle of south Central Asia (Fig. 12), as well as from the looted BMAC burials 
in Afghanistan (Sarianidi 1993: 17, fig. 24). 

The graves contain wheel-made and occasionally hand-made vessels, analog- 
ous to the ceramics found on the settlements. The burial in Khodzha-Goib yiel- 
ded a clay vessel having numerous parallels in southern Bactria (Gotzelt et al. 
1998: 118, fig. 2: 3). In three graves at Tandyryul and Kangurt-Tut there were 
found vessels of the Bishkent culture (PI. 10), in Nurek there was found a knife 
with a curled blade. 

To the circle of farming cultures belong the finds from Parkhar of a vase and 
two silver children's bracelets ornamented with rosettes (Fig. 114: 13, 14) 
Gotzelt et al. 1998: 132, fig. 4: 3; 5); a gold pendant ring with cones in Nurek 
(Fig. 114: 3), lazuli and other stone beads and miniature bronze votive articles; in 
Kangurt-Tut: an ax -adze, a spear, a knife, a razor, a pitchfork, a small cup, a 
mirror, and an earring. Their analogies are known in southern Uzbekistan 
(Ionesov 1990). 

To the farming culture may be assigned chance discoveries from southern 
Tadzhikistan (Fig. 1 14: 16, 18-20). The copper axes from Sangvor and Arakchin 
belong to the Iranian type and have analogies in India in the late Harappan layer 
at Chanhu-Daro, the cemetery of Shahi-Tump and in layer III of the settlement 
of Mundigak; they date back to the late 3rd-early 2nd millennium BC (Kuz'mina 
1966: 8-9, pi. I: 1,2). The ax-adzes were dated to the same time (Kuz'mina 1966: 
14-15, pi. I: 7, 8). The type with a protruding sleeve is represented in Sarazm and 
Dashly 3 (Isakov 1991, fig. 78; Sarianidi 1977: fig. 32) (a late miniature imita- 
tion is present in Kangurt-Tut). From Kara-Pichok in the Hissar valley comes a 
bronze mirror with a handle, typical of the BMAC (Fig. 114: 5). In the Dangar 
region stone weights were found, near Kulyab — two stone batons. In Nurek there 
was found the neck of a silver vessel (Fig. 1 14: 1) whose analogies are present in 
the Fullol hoard in Afghanistan (Tozi and Wardak 1972). These findings indicate 
cultural ties with the farmers of the BMAC culture and Iran and Afghanistan 
going back to the early 2nd millennium BC. The ties with Iran in the late 2nd- 
early 1st millennium BC are demonstrated by the dagger with an incrusted 
handle from Ramit (Fig. 114: 10) Kuz'mina 1966: 52, 53, pi. vii, 9; Pogrebova 
and Chlenova 1970) and the bronze vessel with a spout from Fatmev (Lukonin 



CULTURES OF NORTHERN BACTRIA IN THE LATE BRONZE AGE 275 

1977: 40, fig. 50), whose date — 10th— 8th centuries BC — is determined by the 
discovery of clay ornamented vessels in Talysh, Luristan and the cemetery of 
SialkB (Aliev 1960, pi. lxxx; lxxxiii; ci, ciii). 

The establishment of early ties between the farming regions of Iran and 
Afghanistan with southern Tadzhikistan may have been accounted for by its 
abundant common salt reserves. 

Analysis of the burial rite, ceramics and accompanying articles of Tadzhik- 
istan complexes enables one to assign them beyond doubt to the Molalli and 
Bustan stages of the BMAC and to suggest the population's migration from the 
west, where this culture formed and its early stages are represented. 

Adaptation to the specific mountainous conditions determined the culture's 
peculiarities which allows us to distinguish southern Tadzhikistan as a separate 
local variant of the BMAC. Its distinctions include: type of small settlements 
with a scattered construction pattern; presence of all three types of dwellings; 
blocking up of the graves not with bricks, but with stones and clay; bogar type of 
farming; a serious role of cattle-breeding; wide distribution of the typically 
Andronovo weapon types (arrows, daggers), working tools (sickles, single- 
bladed tailed knives) and ornaments (trumpet -shaped earrings, large beads) as 
well as their production technology employing two-piece stone molds. Great 
importance for Tadzhikistan in the Late Bronze Age is proven not only by 
cultural but also ethnic contacts with the pastoralists of the Andronovo and 
Bishkent cultures which is indicated by the discovery of burials and ceramics of 
these cultures on the BMAC sites and of imported wheel-made pottery on the 
pastoralist's ones (PL 10). There is no doubt as to the peaceful character of the 
interaction between the representatives of the different cultures. There are no 
signs of violent destruction of the farmers' sites, and it is not only material 
culture articles that are widespread but also those of mutual borrowings pertain- 
ing to cults and peculiarities of burial rite. Such symbiosis reflects a most active 
process of ethnic assimilation and is possible only under the conditions of the 
adoption of a new language and general bilingualism. 

Bishkent-Vaksh culture 

The culture bears the names of the two valleys where sites have been 
discovered, the Bishkent by A. Mandel'shtam and the Vakhsh by B. Litvinsky. 
In actual fact, the sites represent two variants of a single culture. The Bishkent - 
Vakhsh culture is the most outstanding cultural formation of Central Asia, 
which reflects the synthesis of the cattle-breeding and farming population in 
the Bronze Age. The monuments are located in the Bishkent valley on the 
Kafirnigan river and in the valleys of the rivers Vakhsh and Kyzylsu. These 
areas are very suitable for cattle-raising. 

The culture is known from a single settlement, Tashguzor (Fig. 91: 1-22), 
and numerous cemeteries: on the Kafirnigan - Tulkhar, Aruktau, Bishkent I, II, 
III, Isanbay, burials at Shakh; on the Vakhsh - Vakhsh, Tigrovaya Balka, 
Oykul', Dzharkul', Amu -Darya, Kyzlar-Kala; on the Kyzylsu - Makonimor 
Ittifok, Guliston, Gelot, Maidapatta, Obkukh, three burials from the BMAC 
cemetery of Kangurt-Tut, the burials of Shulyupu, Teguzak as well as Karim- 
Berdy in the Hissar valley, on the outskirts of Dushanbe, at Hissar, etc; in 
northern Afghanistan on one of the mounds (tepes) of the settlement of 



276 CHAPTER TWENTY-ONE 

Shortughai, in the upper layer, burials of the Bishkent culture have been 
discovered (Mandel'shtam 1966; 1968; Litvinsky 1964; 1973; Mukhitdinov 
1971; Kuz'mina 1972; 1986; P'yankova 1982a,b; 1986; 1989; 1998; Litvinsky, 
Sedov 1983; Francfort 1989; Kuz'mina, Vinogradova 1996; Vinogradova 
1997; 2000; Gotzelt et al. 1998). 

According to the stratigraphic information of the Tulkhar cemetery, the 
settlement evidence of Tashguzor and the evidence of ceramics three stages 
can be distinguished: 1) Andronovo-Fedorovo; 2) early Bishkent; 3) late Bish- 
kent marked by the presence of ceramics with the applied-roller. At Tashguzor 
(Fig. 91: A) this layer is overlain with one containing Kobadian I ceramics 
from the Achaemenid period. 

At the settlement there were found: 1) black-fired hand-made kitchen 
vessels; 2) ceramics of the Bishkent type with white slip, including that with an 
applied-roller on the shoulder of the pot; in small numbers there is 3) wheel- 
made ceramics of the BMAC of the Mollali-Bustan stages, among the articles 
was the leg of a vase of the Mollali type; 4) sherds of Andronovo type. 

The burial rites of the Bishkent culture are extremely diverse. The most 
ancient of them (type I) was found in the cemeteries of Tulkhar and Babashov. 
These comprised kurgan and stone constructions over an earthen pit; the 
deceased were cremated. The graves contained hand-made ceramics. This is a 
typically Andronovo rite of the Fedorovo type. A peculiarity of these burials 
was small stone settings in the figure of a swastika or a circle with a cross. 

On the Kafirnigan the typically Andronovo elements are: round stone 
enclosures (Aruktau, Bishkent III), square stone cairns (Tulkhar, Aruktau), 
burials in stone cists (Tulkhar, Bishkent II), and stone-filled graves (Aruktau). 

The survival of the cremation rite is demonstrated by burials in a pit in 
keeping with the inhumation rite, but with ashes and pieces of coal around the 
skulls. 

Type II in Tulkhar is represented by burials in the flexed position on the 
side in a pit with a downward passage or dromos, that had been blocked up 
with stones. A fire might be made in the entrance and ashes and pieces of coal 
deposited in the chamber. In the graves round hearths were constructed for 
female burials and rectangular hearths for males. There are also dismembered 
burials in small cists. By the deceased were hand-made ceramics of the 
Bishkent type accounting for 70% of all the ceramics, and imported wheel- 
made ceramics of the BMAC accounting for the remaining 30%. 

The latest - III type - is represented by catacombs with the entrances 
blocked up with stones; these contained flexed burials and hand-made ceramics 
of the Bishkent type. 

In the various types of graves were the bones of sheep or ram, metal 
articles, and rarely ornaments, stone arrowheads and beads. 

The type III burials in Kafirnigan are fully analogous to the cemeteries on 
the Vakhsh and Kyzylsu. The burial rite and ceramic complex here was more 
standardized. The burial grounds have from 10 up to 130 constructions. These 
are round, rarely, square, burial mounds with a stone circle (sometimes - 
double) measuring 2-1 5m in diameter. Inside there is a catacomb, the entrance 
to which is blocked up with stones. The dead lie flexed, men on the right, 
women on the left side. There are double burials of a man and woman. The rite 



CULTURES OF NORTHERN BACTRIA IN THE LATE BRONZE AGE 277 

of the ritual fire is evident, sometimes forming a ring around the grave. A large 
number of constructions are cenotaphs. Inside the catacomb there are usually 
placed the ribs of a sheep or ram, rarely metal articles, stone maces and 
arrowheads. 

A distinctive feature of the Bishkent culture is its ceramics. Thirty-percent 
of the pots in the graves are wheel-made BMAC vessels; 70% of the complex 
is the Bishkent type pottery. Most of these vessels imitate the shape of the 
BMAC types but they are hand-made. Several types of ceramics maintain the 
traditions of the Zaman-Baba culture. This complex consists of round-bot- 
tomed kitchen ware, fired black and typically Bactrian light -slip vessels in the 
shape of pots, basins, (shallow) pans, vases and, most importantly, vessels of 
cylindrical and bi-conical shape. The last types survive in Bactria through the 
Achaemenid period and constitute a peculiarity of the Kobadian I type ceram- 
ics (D'yakonov 1953;1954; Kuz'mina 1972). 

The late type of the Bishkent ceramics include vessels with the applied- 
roller, occasionally decorated with notches, which imitate the characteristically 
steppe type of pottery of the Final Bronze Age. 

The economic and cultural type of the Bishkent culture was complex. At 
the settlement there were found imprints of barley grains, pestles, grinders, 
which is indicative of farming. Stockbreeding is reflected by the bones of 
cattle, ovicaprids, the donkey, the horse, and the custom of placing sheep 
bones into the grave. The topography of the sites suggests that the economy 
was based on stockkeeping and in part on mountain bogar farming, i.e., the use 
of land semi-irrigated and only in spring. The population would use mountain 
pastures and was semi-settled. 

The origin of the culture is open to debate. B. Litvinsky and L. P'yankova 
believe that the culture is genetically related to the BMAC and reflects a 
change-over of a part of the farming population to pastoral stockkeeping. A. 
Mandel'shtam and E. Kuz'mina, on the other hand, hold that it was created by 
Andronovo pastoralists and, possibly, representatives of the Zaman-Baba 
culture. They came to use the ceramics of the neighboring farmers and also 
began making hand-made pottery, which imitated in shape that produced on 
the potter's wheel. 

Of decisive importance is the evidence concerning the burial rite. The early 
monuments of the Bishkent culture maintain the characteristic features of the 
Andronovo Fedorovo burial tradition: burial mound, enclosure, stone cist, 
cremation, swastika, and the hand-made ceramics. Later there appeared graves 
with a downward passage and catacombs. The origin of this rite in Central 
Asia remains debatable. It is known both in the Bactria-Margiana culture, but 
its genesis there is unclear, and in the Zaman-Baba culture where it may be a 
heritage of the Catacomb culture of the European steppe. In types II and III of 
the Bishkent burials Andronovo features are preserved: burial mounds and 
stone enclosures, small cists, the position of the deceased and the custom of 
double-burial, the round and rectangular shape of the sacrificial hearths, the 
vivid manifestations of the fire cult. As long as the burial rite is an ethnic 
indicator of a culture, which is upheld even during long-distance migrations to 
another ecological niche, and wheel-made ceramics are quickly borrowed by 
new-comers, there are serious grounds to believe that the creators of the 



278 CHAPTER TWENTY-ONE 

Bishkent culture were by origin Andronovo pastoralists, who came into contact 
with representatives of the BMAC, which is also expressly indicated in the 
farming culture of Tadzhikistan and Uzbekistan. The new Bishkent culture 
arose as a result of this symbiosis. Its advanced stage is represented by the 
cemeteries in the valley of the Vakhsh river, marked by a stable combination 
of cultural characteristics. This culture became one of the components in the 
formation of the cultural complex of Bactria in the Early Iron Age, and this 
was the ceramic complex that continued its development in Bactria in the 
Achaemenid period (Kuz'mina 1971c; 1972a; Sarianidi 1977; Cattena and 
Gardin 1977). Another component was the culture of ceramics with the 
applied-roller of the Final Bronze Age of the Central Asian steppe. The 
Iranian-speaking people of Bactria formed as a result of the interaction of the 
aboriginal population with different groups of new-comers. 

The Andronovo culture in Tadzhikistan 

In northern Bactria, various traces of the Andronovo tribes' presence were iden- 
tified. In the Vakhsh valley near the Kirov sovkhoz a settlement measuring 400 
sq min area was discovered (Fig. 83; Litvinsky and Solov'ev 1972: 41-47). Two 
thin layers were found separated by a sterile layer 0.5m thick. 

In the lower layer, round hearths were uncovered, 0.5m in diameter, surroun- 
ded by a stone ring, around which animal bones and ceramics were concentrated. 
The vessels are gray, of good quality, often polished. Predominant are the typic- 
ally Fedorovo pots with a rounded shoulder and everted rim. Twelve percent of 
the vessels are ornamented, 60% of the ornament being executed in indented 
stamp. Some pots bear ornament over three zones with the zones of the rim and 
neck united which is generally typical of Central Asia ceramics of the Fedorovo 
type and the Tazabagyab culture. There occur pennants and isosceles triangles 
but absolutely predominant are elongated oblique triangles (Fig. 13 AI, 1; II, 2, 
5; III, 1), which constitute a distinctly Fedorovo ware in Tadzhikistan. One pot 
(Litvinsky et al. 1972: fig. 19 B 2) with swollen sides and a short narrow neck 
ornamented with oblique opposite triangles, finds analogies in the Tazabagyab 
culture (Itina 1977, fig. 20: 2,3; 65) 

The layer also yielded stone implements and fragments of casting molds 
indicative of metalworking. 

Determination of an analogy for the vessel from the Vakhsh, settlement of 
Kokcha 15, is of decisive importance, since imported Namazga VI ceramics was 
found at this site, which sets the complex's date as the 15th— 13th centuries BC 
(Itina 1977: 145) and makes it the earliest Fedorovo complex of Tadzhikistan to 
date. 

In the site's upper layer, clay pottery was found with an admixture of gravel, 
half of it was fired red-and-yellow to a fabric template. The shapes of the pots 
cannot be restored. Half of the vessels are ornamented in indented stamp in the 
form of elongated triangles and pennants thus reflecting the genetic succession 
with the previous layer. The other half are decorated in plain stamp in the form 
of zigzags, herring-bone and isosceles triangles (Litvinsky et al. 1972, fig. 20A). 
The date of the upper layer is determined by the finding of a yellow clay vessel 
with a spout (Litvinsky et al. 1972: 14, fig. 19A, 1), which finds analogies in the 
ceramics of the late sites in the Kara Kum and other sites of the turn of the 2nd- 



CULTURES OF NORTHERN BACTRIA IN THE LATE BRONZE AGE 279 

early 1st millennium BC (Litvinsky et al. 1962: 255). Of the same date is the 
stone mold (Litvinsky 1972: 45, fig. 21: 1), designed probably for casting a 
chisel with a wedge-shaped razor and a roller round the sleeve, peculiar to the 
Final Bronze complexes of Kazakhstan (Chernikov 1960 fig. lxiii, 1). 

Andronovo ceramics with comb-made ornamentation was found in the Hissar 
valley at Pakhtabad. Fragments of Andronovo pottery were collected by A. 
Vinogradov in northern Afghanistan on the left bank of the Amu -Darya (Fig. 50: 
54-57). It is also present at Shortughai (Fig. 52: 24-33). In addition, sherds of 
Andronovo-type ceramics were discovered by A. Ranov on the left bank of the 
Vakhsh river at the settlement of Karabura (Litvinsky 1972: 46) and at the Dzhi- 
likul' crossing (P'yankova 1998: 17 o). On top of that, Andronovo ceramics and 
bronze articles were found in the Vakhsh 's upper reaches near Nurek in the up- 
per layer of the settlements of Tutkaul and Sayed, belonging to the Hissar culture 
(Ranov and Korobkova 1971: 146; Yusupov 1975: 139). An Andronovo burial 
was discovered in the Nurek area at the Sharshar pass near the town of Dangara 
at the ancient cemetery at Kirov (P'yankova 1999: 288). Andronovo sherds were 
identified in a layer of the Bishkent culture at the settlement of Tashguzor 
(P'yankova 1999, fig. 10). In the south of Tadzhikistan at the confluence of the 
Kyzylsu and Pyandzh rivers at the ancient town of Saksanokhur there has been 
found a hand-made vessel and a wheel-made pot of the Biskent culture (Fig. 83: 
19; Litvinsky and Mukhitdinov 1969: 169, 161; fig. 1, 2). The hand-made vessel 
has a widely swollen body and a narrow neck. Underneath the neck there are 
oblique hanging triangles applied in indented stamp, below along the body there 
went a strip of hatched zigzags. Analogies to the ornament and shape of the 
vessel are known in Khorezm in the cemetery of Tagisken (Itina and Yablonsky 
2001: 94-101; fig. 82: 507; 83: 517: 94) dating from the 9th-8th centuries BC. 

To the Andronovo culture also belong three cemeteries of Tadzhikistan: Da- 
shti-Kozi (Isakov and Potemkina 1989; Bostongukhar 1998), Kumsay (Vinogra- 
dova and P'yankova 1990; Vinogradova 2000) and Tuyun (Vinogradova 2000). 

The Dashti-Kozi cemetery is located in the foothills of the Turkestan ridge. 
In the upper reaches of the Zeravshan river, on its right bank, 27 graves were ex- 
plored. It is difficult to judge their structure because according to the field diaries 
and drawings the outlines of the grave pits and dromoi were not established 
which challenges the credibility of their publication (Fig. 48: 4-14, 20, 22-23, 26; 
Bostongukhar 1998). Moreover, though it was only in burials 25 and 25 (1) that 
catacombs were identified, the publication (Isakov and Potemkin 1989: 145) re- 
constructs all the burial structures as catacombs. The entrance to the shallow en- 
trance pit is blocked up with stones. Prevalent are single burials strongly flexed 
on the left side, head to the west with deviations. There were also identified: a 
burial of a mother and child, paired heterosexual burials of various dates, and 
one triple burial, dismembered backbones and a child's skull. Before the burial 
in the entrance pit one would make a fire and deposit ocher; a vessel was placed 
by the deceased, women have ornaments, several graves contain sheep bones. In 
catacomb 25 there was revealed the simultaneous burial of an adult male and 
female with eastward orientation of the head. The dead are lightly flexed, some 
of them have one arm extended. There are accompanying articles. T. Khodzha- 
yov has established that three women and two men belong to the southern 
Mediterranean type, two men, distinguished by a great height — to the 



280 CHAPTER TWENTY-ONE 

Andronovo type. Another male Andronovan is buried with his head to the east 
along with two aboriginal women in grave 12 without accompanying articles but 
with sheep bones. The rite's distinctions in grave 25 reflect the ritual character of 
the complex, or the fact that the dead were either foreigners or victims of an 
epidemic. Apart from this, in the cemetery there were discovered shallow ritual 
pits with traces of repeated fires and oil poured on (Bostongukhar 1998: 84 - 87). 

In the female burials there were found bronze or gold plaques for plaits, 
trumpet-shaped earrings, bracelets, mirrors, large ribbed bronze beads, lazuli 
beads, bronze beads used as decorations on red boots. Three rich burials were 
identified ## 3, 15 and 25 (1), that contained women wearing 3-4 bracelets on 
their wrists, gold temple rings, earrings with a hook, a wreath of beads, a mirror 
with a loop, a pendant, and dresses decorated with beads (Bostongukhar 1998: 
69, 74, 75, 83, fig. 36, 45). The bronze has a considerable content of tin and lead. 

The ceramics are hand-made of gray and yellow-and-orange color executed 
in clay with an admixture of fired clay and large gravel. The principle form is a 
wide-mouthed pot with an accentuated neck and a rounded shoulder. The body's 
maximum width is 2/3 that of the height, for some pots it is at mid height which 
is a late feature of the Fedorovo pottery. Some pots are of elongated proportions 
and have a narrow mouth making them close to jugs characteristic of the Final 
Bronze Age. Half of the pottery is ornamented in plain or combed stamp, the 
ornament being located in one zone over the shoulder: hatched isosceles triangles 
with an upward — or downward-pointed apex, vertical or horizontal herring- 
bone — on two vessels it goes over the rim and shoulder; zigzags going down 
from the rim; one pot has flutes over the neck which is typical of Fedorovo 
ceramics, but the proportions of the vessel are elongated (Bostongukhar 1998: 
figs. 2, 3: 7; 25: 8; 26: 2; 28: 11; 30: 10; 31: 26, 27; 32: 24; 36: 6). 

Contrary to the view of A. Isakov and T. Potemkina, this pottery neither in 
terms of technology nor of types has any similarity with the late Alakul' 
ceramics but belongs to the late degraded Fedorovo tradition, finding its closest 
analogies among the vessels of Tadzhikistan cemeteries. 

Of interest is the material recovered from burial 7: this paired burial yielded 
an Andronovo pot, a Molalli-type pot and a similarly-shaped one (Bostongukhar 
1998: fig. 28: 10; 27: 1), finding analogies in the wheel-made pottery of 
Tadzhikistan, in particular, in the cemetery of Kumsay (Vinogradova and 
P'yankova 1990: fig. 5: 2, 3). The narrow-mouthed pot from burial 9 
(Bostongukhar 1998: fig. 30, 11) has a typological similarity with the ceramics 
of the Late Bronze settlement of Yakke-Pareang (Itina 1977: fig. 82: 24; 25) and 
the mausoleum in Tagiskena (Itina et al. 2001, fig. 87: 570, 573; 88: 574) in 
Khorezm. 

Another Andronovo cemetery of the Fedorovo circle is Kumsay, which was 
discovered in the piedmont territory of the Hissar valley (Fig. 49: 1-3, 15-18, 24, 
27; Vinogradova and P'yankova 1990: 98-112; Vinogradova 2000: 90, fig. 4). 
Twenty-one catacomb-type graves were excavated; the top grave was marked 
with a few stones, the entrance to a catacomb was blocked up with stones. The 
burials are strongly flexed, head to the north-east, women on the left side, the 
adolescent on the right. By the head there is a vessel; women wear ornaments. 

Six graves yielded hand-made clay ceramics with an admixture of lime and 
clay fired gray. There one finds pots with an open mouth, a rounded shoulder, 



CULTURES OF NORTHERN BACTRIA IN THE LATE BRONZE AGE 28 1 

occasionally, wide sides or of elongated proportions. Two vessels are decorated 
with flutes over the shoulder; the rest are ornamented in indented stamp in two 
zones — over the rim or neck and shoulder in the form of two rows of horizontal 
zigzags or opposite triangles (Vinogradova et ah 1990, fig. 5: 4-7). This is the 
pottery of the late Fedorovo tradition. 

In the five children's graves there were found red-and-orange vessels with 
white slip made on the potter's wheel. These are two pots, an open basin, and a 
narrow-mouthed jug (Vinogradova et ah 1990: fig. 4: 12; 5: 1-3). This pottery 
finds analogies in the farming Molalli stage cemeteries of southern Tadzhikistan 
(P'yankova 1989a, fig. on p. 169 I). 

Women's ornaments are represented by temple pendants with a hook and one 
spectacles-like pendant, wide, open bracelets, large bronze beads and polytypic 
paste beads and large bronze ribbed beads. They were used in the embroidery of 
boots. According to I, Ravich's conclusion, the decorations contain 10% tin and 
5-10% lead. Typically and in terms of their composition and casting technique, 
they are characteristic of the late Andronovo complexes of the Zeravshan and 
Tadzhikistan (Vinogradova et ah 1990: 109). 

In the basin of the Kyzylsu river the Tuyun cemetery has been discovered 
(Fig. 71: 10, 11; Vinogradova 2000: 105, fig. 19). The burial of an adolescent 
girl was made in a catacomb in the flexed position on the left side; nearby there 
are two children's skulls and a Fedorovo hand-made vessel with a rounded 
shoulder, ornamented in indented stamp with pendants over the rim and neck. 
There were found a silver pendant with a hook, paste beads on the skull and 
large bronze ones on the legs. 

N. Vinogradova and L. P'yankova (1990: 110; 111; P'yankova 1998: 70; 
Vinogradova 2000: 105) assign the cemeteries of Tadzhikistan to the Zeravshan 
type of the Andronovo -Tazabagyab community singled out by N. Avanesova 
(1985: 40) and they suggest a migration of this population from the Zeravshan. 
A. Isakov and T. Potemkina (1989: 165) view southern Kazakhstan as the 
original territory. Both suggestions are wrong. The southern Kazakhstan sites of 
Tautary and Kuyukty have nothing in common with those of Tadzhikistan (a 
fence, a flat grave, cremation, richly ornamented pottery including articles with a 
ledge on the shoulder). The differentiation of the Zeravshan variant does not 
appear convincing either, since the burial rite and ceramics of the mid Zeravshan 
sites are very diverse and reflect the presence along with the Fedorovo traditions 
of Alakul' and Timber -grave ones as well as of the BMAC farming culture (the 
Dzham hoard; Fig 84: 1-11). One can only assert the typological similarity of 
metal alloys and the set of ornaments represented on the mid Zeravshan and in 
the cemeteries of Tadzhikistan: Fedorovo trumpet-shaped earrings, wide open 
bracelets, large beads with a rib, and stone beads. But ornaments do not 
constitute a culture-determining characteristic. Of decisive importance are the 
ceramic complexes. As for the burial rite in a catacomb, this reflects the 
influence of the BMAC culture. As long as the center of this culture is situated in 
southern Tadzhikistan and Uzbekistan, the formation of Tadzhikistan' s peculiar 
cemeteries is probably to be situated in the west where the oldest Andronovo 
settlement is located at the Kirov sovkhoz (lower layer) and where Andronovo 
features are prominently represented in the cemeteries of Tulkhar and Babashov. 
From there also originate the evolution of the catacomb, the blocking up of the 



282 CHAPTER TWENTY-ONE 

entrance with stones, the custom of making fires in the entrance and placing 
ashes in the chamber. It is precisely the Bishkent culture that is characterized by 
a combination of imported BMAC pottery and local hand-made ceramics. But in 
Bishkent the synthesis of the Andronovo and farming traditions led to the 
formation of a new synthetic culture, represented by the Vakhsh cemeteries. 

How old are the Andronovo sites? The most ancient evidence is the discovery 
of an Andronovo sherd in the post-Harappan layer at Shortughai. At the Vakhsh 
site a lower layer was identified but its ceramic complex is analogous with the 
pottery of the majority of other sites. Fedorovo materials were found on the 
BMAC sites at the settlements and in the cemeteries of Dzharkutan, Bustan 6, 
Tandyryul, Teguzak, Kangurt-Tut, Shortughai and at the Bishkent culture 
settlement of Teguzor (PL 10), which allows one to synchronize these cultures. 
The chronology is determined by metal articles. 

The dagger with a butted blade finds numerous analogies all over Kazakhstan 
including in the complexes of the settlements with the applied-roller type of cer- 
amics: Alekseevka, Chelkar, Elizavetinskiy priisk, Stalinskiy rudnik (Fig. 53: 24, 
25, 30), Atasu, Il'inka, Kent (Krivtsova-Grakova 1948, fig. 20; Orazbaev 1958: 
pi. viii,13; ix,4-6; Zdanovich S. 1979; Margulan 1979, fig. 136,3; Matyushchen- 
ko 1973, fig. 2, 5; Gening and Eshchenko 1973, fig. 4, 10; Varfolomeev 1987, 
fig. 6, 27; Kadyrbaev et al. 1992: fig. 20: 3) as well as the late Timber-grave 
complexes of the Volga region and Ukraine as far as the Danube region (types 
34, 36 according to E. Chernykh 1976), including the following hoards: Malye 
Kopani, Golovurovo, Loboykovo, Kabakovo, Sokoleny, Voloshskoe, Krasny 
Mayak, which indicate that the Central Asian examples date to the 13th— 10th 
centuries BC. 

The daggers of Tulkhar, Vakhshuvar and Dashly 3 (Fig. 53: 1, 6, 10) are 
characteristic of eastern Kazakhstan (Chernikov 1960: 164, pi. lxiv, 2-4); another 
variant is present at the settlement of Novonikol'skoye (Zdanovich 1988, pi. 10, 
21). Variants comparable with the single-bladed knives with a bent back blade of 
Kangurt-Tut and Nurek (Fig. 53: 11, 31) are known in the Urals in the Vese- 
lovskiy hoard together with an Andronovo-type axe, a chisel, a spear and a sickle 
(Obydennkov 1996: 109; fig. 3: 18) and in Semirech'e. The Nurek type of knife 
is represented at the central Kazakhstan settlements of Karkaraly and Kent along 
with applied-roller ceramics (Margulan 1998, fig. 134: 7; Varfolomeev 1988, fig. 
3). Various types of sleeved arrows with a protruding or hidden socket pertain to 
the category of weapons that had developed in the Andronovo culture since the 
Sintashta period and right to the Late Bronze Age, when variants analogous with 
those from Central Asia are represented in the complexes with the applied-roller 
ceramics all over the Andronovo range: Alekseevka, Sargary, Shortandy-Bulak, 
Kent, Malokrasnoyarka and others (Krivtsova-Grakova 1948, fig. 33, 1; Zdano- 
vich 1988, pi. 10, 8-10; Varfolomeev 1988, fig. 4, 27; Margulan 1988: fig. 127: 
8-10; 190). For the most part, these articles date back to the 13-1 lth centuries 
BC, several types survived till the 10-9th centuries BC. The attribution of the 
monuments to the Final Bronze Age is borne out by the recovery at the settle- 
ment of Dzharkutan of a cylindrical seal (Meytarchiyan 1984), belonging to the 
same style as the seal with a rider from Merv (Kuz'mina 1971) and a number of 
seals from Southern Bactria, published but not dated by V. Sarianidi (1998c). 



CULTURES OF NORTHERN BACTRIA IN THE LATE BRONZE AGE 283 

Another basis for determining the age of the Andronovo sites in northern 
Bactria is the ceramics. The shape of most of the vessels with a rounded shoulder 
and a wide and tall mouth, the ornament in indented stamp by oblique net in the 
form of horizontal zigzag, herring-bone, oblique triangles, pennants and meander 
are typical of the Fedorovo tradition. But there also occur vessels of elongated 
proportions or with a globular body or a low mouth. The main differences from 
the classic Fedorovo pottery are found in the ornamentation. A large percentage 
of pots are altogether devoid of ornament, the ornament is simplified and applied 
over one zone — mouth or shoulder, often different zones bear the same motif. 
These irregularities are peculiar to the late Fedorovo ceramics of the whole 
Andronovo range. Particularly typical of the late Fedorovo pottery is the pattern 
motif of very elongated, often, opposite oblique triangles and wisps of oblique 
lines. They are known on the late Fedorovo ceramics of the Urals and the Tobol, 
often co-occurring there with the applied-roller type of ceramics and the pottery 
of the Cherkaskul' and Lugavskaya cultures which is absolutely uncharacteristic 
of Central Asia (Obydennov 1977: fig. 10; Obydennov and Shorin 1995: fig. 39, 
46; Stefanov and Korochkova 2000: fig. 27). 

Similar ornaments are represented on the pottery of the late Andronovo sites 
of central Kazakhstan: Egizek, Belasar (Margulan 1998: fig. 32; 37), Ikpen' III, 
Shapat (Tkachev 2002 part 1: fig. 36; part 2: 180) and eastern Kazakhstan: 
Izmaylovka (Akishev 1987: fig. 29; 47), where these ceramics were found along 
with the applied-roller pottery of the Dongal type and the Begazy culture and are 
dated to the pre-Saka epoch of the 9th-8th centuries BC according to the 
discovery of a wedge-shaped chisel, arrows and the most ancient bits and cheek- 
pieces. Also typical of Izmaylovka is the construction of stone cists, and the 
presence in the grave of a funerary fire and ocher. 

Vessels with a spout recorded for northern Bactria in the latest complexes 
were found in central Kazakhstan at the settlement of Tagibay-Bulak (Margulan 
1998: fig. 140: 1) and in the cemetery of Krasnye Gory (Tkachev 2002, 2: fig. 
195) together with ceramics with applied -roller decoration of the Dongal type 
and these are dated to the 10th 9th centuries BC by the discovery in Krasnye 
Gory of two-ring iron bits. 

This date serves as the upper limit for the Bronze Age sites of northern 
Bactria. The main part of the BMAC and Andronovo sites containing the 
Alekseevka type pottery with applied-roller decoration and bronze articles 
characteristic of the Eurasian steppes from the Danube to China should be placed 
back in the 13-1 1th centuries BC. 

The time of the first arrival of the Fedorovo tribes in the south of Central 
Asia is not known. Possibly, they advanced in several consecutive waves. But 
judging by the multitude of sites (over 25) the peak of their migrational activity 
fell within the Late Bronze Age — the time of numerous and multi-directional 
movements all throughout Eurasia's steppes. 

The question whence the Andronovo tribes came to northern Bactria remains 
open. Most likely, it was from the north or north-east, as is suggested by the 
prevalent types of the burial rites: round and rectangular stone enclosures, cists 
and ground pits, co-occurrence of cremation and inhumation, the Fedorovo type 
of pottery with a rounded shoulder but poorly ornamented, often located in one 
or two zones, the abundance of unornamented pottery, and finally, the fact that 



284 CHAPTER TWENTY-ONE 

the Bactrian metallurgical center undoubtedly belongs to the eastern Andronovo 
metallurgical province characterized by distinctive types of knives, daggers, 
celts, and particularly, ornaments: trumpet-shaped earrings, horned bracelets, 
large cast beads as well as alloys with a high tin and, often, lead content. 

The relations between the aboriginal farmers and the pastoral newcomers 
were amicable. The peaceful co-existence must have been promoted by 
economic and cultural specialization: the Andronovans were semi-settled cattle- 
breeders and suppliers of bronze articles in exchange for which they would 
receive farming produce, stone beads and wheel-made pottery. 

In the north — among the Andronovo sites proper — traces of interaction with 
the farmers have been identified: discoveries are known of stone beads, 
especially of lazuli and turquoise, executed by the jewelers of Namazga VI- 
BMAC on sites of the steppe zone of Central Asia (Gurdush cemetery, sites at 
Makhan-Darya, etc.) (Gulyamov et al. 1966, p. 202, pi xxiii, xxiv); similar stone 
beads now and then occur from the Urals to Siberia at the sites of Alabuga, 
Ushkatta, Keembay, Ural-Saj, Aksayman, Borovoe, Nurtai, Rostovka, Sopka 2; 
common at the Kozhumberdy and Atasu types of Andronovo sites are plaques 
with complex patterns, imitating bronze and stone seals of the farmers of 
Margiana and Bactria of the Namazga VI period (Margulan 1998: fig. 191; 
Tkachev part 2: fig. 137, 167); a Namazga-type pin with a bi-spiral head was 
found in Khorezm (casting mold from Kokcha 15a), in northern Kazakhstan 
(imitation from the Borovoe cemetery) (Kuz'mina 1988c). However, the number 
of imports in the north in the 15th 14th centuries BC is so negligible that it 
points to the sporadic character of contacts with the farmers. It was only in the 
13th 1 1th centuries BC that the pastoral tribes began to occupy extensively 
Central Asia's deserts and high mountains. 

The 13th— 9th centuries BC mark the transition to the mountain pasture type 
of cattle-breeding, the mastery of riding skills, the appearance of iron and other 
cultural innovations as well as active advancement to the south (Map 10). The 
applied-roller ceramics, often with 'moustaches', spreads in the steppe (Fig. 51). 
It is typical of the Noua culture in Romania, the late Timber-grave Sabatinovo 
and Belozerka stages in the Ukraine, the IV-Khvalynsk (Ivanovo) stage in the 
Volga region, the late Andronovo monuments of the Alekseevka type, represen- 
ted in the Urals, all over Kazakhstan, the Altai region and Semirech'e as far as 
Mongolia (Krivtsova-Grakova 1948; Terenozhkin 1961; Margulan et al. 1966; 
Margulan 1979; Leskov 1970; Orazbaev 1958; Zdanovich S. 1979; Potemkina 
1979; Chernikov 1960; Mogil'nikov 1976; Bernsham 1952; Chernykh 1983; 
1992; Kuz'mina 1964a; Tsybiktarov 1998). It appears also in Central Asia. 

It was specifically in the Late Bronze Age that contacts between the farmers 
and the steppe intensified. 

In the cemetery of Kulevchi VI in the southern Urals, two cremation burials 
were excavated. The bottom of the pits was heavily burnt, above it was a clay 
layer followed by that of charcoal. In each grave were discovered the calcined 
bones of humans and animals, sheep ribs and two vessels apiece: one was a 
richly ornamented Fedorovo vessel, the other a large basin imitating wheel-made 
pottery (Vinogradov 2000: 30, 31, fig. 9). Wheel-made pottery was identified at 
the settlement of Chernoozer'e on the Irtysh (Viktorova et al. 1974: 22, 23, fig. 
2: 2). Pottery of the late Namazga VI-Molalli and Bustan stages made on the 



CULTURES OF NORTHERN BACTRIA IN THE LATE BRONZE AGE 285 

potter's wheel or imitating it in shape was found at the late Fedorovo settlement 
of Pavlovka (Fig. 72) in northern Kazakhstan (Malyutina 1985; 1991), in the late 
Tautary cemetery in southern Kazakhstan (Maksimova 1962), at the settlement 
of Bien in Semirech'e (Karabaspakova 1987: 96) and in central Kazakhstan at 
the settlements of Kent and Myrzhik and in the Tasyrbay cemetery (Varfolomeev 
1991: 15, 17) in combination with Alekseevka ceramics. The co-occurrence in 
Kent of a large number of articles that can be reliably dated by the European 
scale to the 13th 10th centuries BC provides grounds for establishing the chron- 
ology of Namazga VI and the BMAC. Wheel-made pottery along with Alekse- 
evka ceramics was recorded also at a number of settlements in the Altai region: 
Kalinovka, Kureika 3, Pereezd, Burla 3, Molokovo II, Chekanovskiy Log 
(Ivanov 1987; 1989; Udodov 1988; Kiryushin and Luzin 1990; Kiryushin et al. 
1990, p. 117, 122; Udodov 1991, 75; Tishkin 1998: 92; Demin and Sitnikov 
1998: 94). Noteworthy is the fact that it is present only in the regions of ore 
deposits along the trade routes. 

Andronovo sites of the Pamirs 

The Andronovo late Fedorovo tribes also occupied the high mountain territory of 
the eastern Pamirs. Here have been discovered the Mashale rock shelter, the 
Kurteke cave, the cemeteries of Karadimur, Kokuybel'su, Yuzhbok and the late 
cemeteries of Kyzyl -Rabat and Zhartygumbes (Fig. 83: 20; Litvinsky 1972: 29, 
30, 90; Litvinsky and Ranov 1964; Ranov 1962: 7, 8; Babaev 1980: 75-87; 1984: 
104-106; 1989; P'yankova 1998: 154). In the Karadimur cemetery a complex 
square construction was revealed with double stone walls filled in with clay and 
an entrance -ramp. In the center there is a grave oriented to the south-east and 
covered with stones. Inside the grave there is a pot and fragments of several 
more vessels. The Karadimur construction resembles the mausoleums of the 
Dandybay culture. In the other graves the burials of cremation and inhumation 
are set in primitive stone constructions and accompanied with vessels. 

The ceramics are coarse. There are jars and pots with a rounded shoulder, 
unornamented or with poor decoration applied over the shoulder in the form of a 
meander in coarse -indented and plain stamp. 

In terms of the burial rite and ceramics, the Pamir sites are reminiscent of the 
late Fedorovo traditions and typologically are in part comparable with the pre- 
Saka complexes of Semirech'e. They date from the 11th to the 9th centuries BC. 
The few bronze articles are made of the high-tin Andronovo bronze. The spear 
from Yuzhbok is forged from arsenical bronze. The topography of the cemeteries 
points to seasonal cattle drives. The Kurteke cave, where an arrowhead was 
found, served as a shelter for hunters. 

The Pamirs have also yielded petroglyphs, among which are representations 
of chariots: Tekketash and Akdzhilga (Mandel' shtam 1961; Zhukov and Ranov 
1972; 1974). 

For the late Andronovo population of the Pamirs the way to India was open 
through the passes into the Indus valley. 



286 CHAPTER TWENTY-ONE 

The Andronovo culture in Southern Bactria 

Andronovo monuments were discovered by the Soviet-Afghan expedition. 
Andronovo ceramics are represented in the oasis of Dashly (Sarianidi 1977, fig. 
66) and on the southern bank of the Amu -Darya in A. Vinogradov's collections 
(fig. 50: 54-57). These comprise sherds of hand-made ware decorated with tri- 
angle imprints and in plain stamp in the form of vertical herring-bone and 
isosceles triangles. The applied-roller ceramics is absent; the likely date is the 
15th 14th centuries BC, though a later date is not ruled out. 

Andronovo materials were discovered at the settlement of Shortughai (Franc- 
fort 1977; 1989, Kuz'mina 1992a). H.-P. Francfort separates the identified con- 
struction horizons into two stages. The first is Harappan. 

The ceramics are wheel-made and includes the specifically Harappan shapes 
and ornamentatal motifs (Francfort 1989: pi. 32-38, 43, 46-47, 52-54, 59-61) as 
well as the seals, beads, and shell artifacts typical of the civilization in the Indus 
valley (Francfort 1989: pi. 80, 81, 1-14, 75). As it appears, the common transport 
types and the uniformity of a range of ceramic forms and metal artifacts testify to 
active ties between the inhabitants of the Harappan colony and the southern far- 
ming population of Central Asia in the first stage of the monument's occupation. 
The common types allow one to synchronize Shortughai (stage A) - Harappa- 
Namazga V-VI - Mundigak IV - Shahr-i-Sokhta III-IV - Yahya IVB - Bampur 
V-VI - Kulli-Shahdad-Hissar IIIC. In advocating a long chronology, H.-P. 
Francfort dates the first stage to 2200-2000 BC when the settlement was aban- 
doned. 

The second stage of the monument reflects a new culture which spread from 
northern Bactria. It is a ceramic complex finding analogies in Dashly and Sap- 
palli. In the upper (IV) stage of the second period there were discovered burials 
of the Bishkent culture containing the type of ceramics of the northern-Bactrian 
BMAC variant, made on the potter's wheel, together with hand-made pottery 
(Francfort 1989: pi. xi, xxv) which is also characteristic of Tadzhikistan cemet- 
eries (P'yankova 1989, fig. 61-63, pi. 77: 15, 17; 78, 4). In addition, Andronovo 
ceramics of the Fedorovo type were found (fig. 52, 24-33). These are hand- 
made, occasionally polished, vessels without ornamentation and fragments deco- 
rated with the ornament peculiar to the late Fedorovo complexes, executed in 
indented and plain stamp on an oblique net (Francfort 1989: 80.242, pi. 58, 1, 3, 
4, 12-15, xxvi, 3). 

A special typological group consists of the hand-made pottery with an 
applied-roller underneath the ridge or ornament imitating such a roller (Francfort 
1989: pi. 58, 6-9, xxvii, 1). Of the material articles, correlating with the steppe 
ceramic complex are the following: a bronze double -bladed arrow with a hidden 
sleeve (Francfort 1989, period IV, pi. 78, 3), and metal artifacts (Francfort 1989: 
208, pi. 75) made of bronze or a compound alloy with arsenic and tin (periods 
III, IV), in contrast to the stage A articles, executed from copper and arsenical 
bronze, which is a hallmark of the metallurgical production of all the farming 
centers of Central Asia and the Zaman-Baba culture. As for the sites of the 
Bishkent culture, they, as already mentioned, are characterized by the same co- 
occurrence of articles of arsenical and tin bronze as in Shortughai; the latter as a 
rule belong to Andronovo types (Kuz'mina 1966: 85-91; Terekhova 1975; 
Ruzanov 1982; P'yankova 1989: 86). According to the conclusion of the paleo- 



CULTURES OF NORTHERN BACTRIA IN THE LATE BRONZE AGE 287 

botanist G. Wilcox (Francfort 1989, 180, pi. 65), in Shortughai's upper layers 
one finds cereals of Northern Eurasian origin: millet and hexastichous barley, 
identified in Kangurt-Tut. 

Of interest is the stratigraphic position of the Andronovo ceramics. One sherd 
was found in the upper (II) period of the Harappan settlement, all the others 
belong to the second stage: some in period III and twice as many in the last — 
period I — of the monument's occupation where they co-exist with the Bishkent 
closed complexes (Figs. 52: 24-33; 91: 23-27). The following succession of cul- 
tures is in evidence: (stage A) = periods I and II = Harappa = Namazga V; stage 
B = period III = northern Bactrian variant of the BMAC and Andronovo; period 

IV = BMAC = Andronovo = Bishkent. H.-P. Francfort, recognizing the continu- 
ity of the settlement's occupation after the departure of the Harappan population, 
which is borne out from his point of view by the preservation of some traditions 
of ceramic manufacture and construction (large-format brick), and advocating 
the long chronology, dates stage B to 2000-1700 BC, using the C 14 calibrated 
dates and emphasizing the incompatibility of his chronological scale with the one 
accepted in our Russian scientific tradition. As for the 16 dates he obtained 
(Francfort 1989, 241, pi. 81), they do not look overly convincing in view of their 
great spread and overlap: period I - 4075-3725 (2850-2090 BC), period II - 
4375-3710 (3340-2060 BC), period III - 3620-3050 (2850-940 BC), period IV - 
3640-3535 (2160-1680 BC). The same is true for the series of dates for the sites 
of the Andronovo and Bishkent cultures (Appendix). Therefore it makes sense to 
refer to the comparative historical analysis and synchronization of the comp- 
lexes. The relative chronology and synchronization of the monuments of Shortu- 
gai's early stage, offered by H.-P. Francfort, are unquestionable. The advocates 
of the long chronology just as well have no doubts as to the lower date - late 3rd 
millennium BC which is well consistent with the beginning of the Namazga V 
stage according to V. Masson's chronological scale (1956) accepted by the majo- 
rity of local scholars. However, I judge that there are no good grounds to assign 
the time of the Harappans' residence in Shortughai only to the beginning of the 
Namazga V period, whose end is determined as 1700-1600 BC, being synch- 
ronous with the most recognized upper date of the existence of the Mohenjo- 
Daro civilization (Bongard-Levin and fl'in 1985, 92-95; Elizarenkova 1989: 432, 
434; Possehl 1990; 1993; 1994; 1999). V. Masson (1959: 109ff), in pointing out 
a crisis of urban civilization in the late Namazga V period, emphasized that in 
the Namazga VI period the life of urban centers went on and crafts reached an 
even higher level. V. I. Sarianidi (1977: 151-5) sees during this period the flo- 
rescence of urban culture both in Margiana and in Bactria — the florescence that 
determined the formation of the Bactrian state in the pre-Achaemenid period. 

The chronological system of the south of Central Asia is based on analysis of 
the stratigraphy of a large number of sites which provided a successive thick for- 
mation of layers bearing evidence to the long-lasting existence of the Namazga 

V and Namazga VI stages (PL IV, V). The upper chronological reference point is 
the Achaemenid period - the 6th century BC. The succession of the manufac- 
turing traditions of the Bactrian white wheel-made ceramics from the Bronze 
Age to the Achaemenids is incontestable (Kuz'mina 1971c; 1972a; 1972b). This 
thesis was reliably corroborated from an independent analysis of a mass of 
ceramic material (Sarianidi 1977: 116; Cattenat and Gardin 1977). This rules out 



288 CHAPTER TWENTY-ONE 

the possibility of dating the end of the Namazga VI period to 1700 BC. The 
chronology of Namazga VI and its Bactrian variant has been refined through the 
discovery of the co-occurring steppe and farming ceramics at the settlements of 
Shortughai, Dzharkutan, Kangurt-Tut, and Teguzak, at the settlements and in the 
cemeteries of the Bishkent and Andronovo cultures: Dashti-Kozi and Kumsay. 
Of crucial importance in determining their chronology as 13th— 1 1th centuries 
BC is the discovery of applied-roller ceramics dated according to the European 
scale. 

Andronovo metal is known at the farming settlements of Bactria and Mar- 
giana: it is represented by a dagger from Dashly 3 and arrows from Togolok 1 5 
and Shortughai as well as the chance finds of knives, daggers, sleeved spears and 
a baton in the form of the upper part of a horse from southern Bactria (Sarianidi 
1977: fig. 36; 1990, pi. xc, 9; Francfort 1989: pi. 78, 3; Amiet 1977: fig. 15, 8; 
Pittman 1984: fig. 32). Of undoubtedly Andronovo provenance is a trumpet- 
shaped earring from northern Afghanistan (Sarianidi 1993: 17, fig. 24). Also 
interesting is a ceremonial bronze axe among the chance finds in the cemeteries 
of southern Bactria (see Ligabue et al. 1988, 164, fig. 101). The blade ends with 
a horse's head with the mane upright (Fig. 95: 5-8). This representation of a 
horse is peculiar to the Andronovo staffs (cf. 55: 6, 7). The northern steppe 
origin of this stylistic manner is indisputably borne out by numerous represen- 
tations on the petroglyphs of Kazakhstan (Pyatkin and Miklashevich 1990, fig. 1, 
2). It was also from there that the horses themselves would be brought to Bactria. 

Metal articles of the types characteristic of the late Andronovo metallurgical 
province spread as far as northern Hindustan: these include a celt from the 
settlement of Ghaligai in Swat (Stacul 1967: fig. 12h) and a celt from 
Kuruksetra — the battlefield of the Aryans, described in the Mahabharata — 
which, in D. Gordon's view (Gordon 1958, 138, fig. 1), was brought by the 
Aryans from their original home. 

The advancement of the pastoral tribes to the south is likely to have 
accounted also for the appearance at the settlement of Pirak in Baluchistan of 
figures of horses and Bactrian camels (Jarrige and Santoni 1979: 177-179, fig. 
94-95; Fig. 104) and the distribution in the north of Hindustan including Thor 
(Fig. 34: 15) of representations of chariots on petroglyphs, stylistically unlike the 
profile of those from the Near East but similar to the Andronovo examples 
designed schematically (Kasambi 1968: 123; Kuz'mina 1980c: 34; Lai 1961, 
figs. 5, 6; Jettmar 1985). 

In the late 2nd millennium BC in the south of Central Asia, in Iran and 
Afghanistan, decoration of pottery with applied-roller ornament became com- 
mon. Vessels with rollers were found in Tepe Giyan (Figs. 51; 57: 5-8) whence 
originate the typically steppe two-ring bits, cheek-pieces and plaques with a loop 
(Contenau and Ghirshman 1935: 8, 13), which Ghirshman at first associated with 
the migration of Iranian horsemen from the steppes. Hand-made ceramics with 
applied-roller decoration is present in Afghanistan at Tillya-Tepe (Sarianidi 
1972, fig. 7.2, 9.1, 12.1, 22.4). Analogous pottery is part of the complexes of the 
1 1 th 9th centuries BC in Turkmenia (Yaz I), Uzbekistan (Kuchuk), Tadzhikistan 
(Karim-Berdy, where a Kazakhstan-type celt was found) (Masson 1959; Askarov 
and Al'baum 1979; Vinogradova and Kuz'mina 1986). The distribution in the 
south of the fashion to ornament pottery with applied-rollers, which is 



CULTURES OF NORTHERN BACTRIA IN THE LATE BRONZE AGE 289 

technologically unjustified on wheel-made ceramics and represents an imitation 
of hand-made vessels, and the appearance of metal articles, belonging, in terms 
of both their type and composition (bronze with a high tin content) to the 
production of the northern metallurgical seats and having no prototypes in the 
Near Eastern cultures, point to the main direction of migrations in the 12th— 9th 
centuries BC which was from north to south, from the territory of the steppes. 

The culture of the Barbarian Occupation Period 

The final of Yaz I-Kuchuk II, the ancient farming culture of northern Bactria of 
the barbarian occupation period, is demonstrated by the settlement of Karim- 
Berdy and the upper layer of Kangurt-Tut and Teguzak (Vinogradova and 
Kuz'mina 1986: 138-146, fig. 9: 4, 5, 11; P'yankova 1996: 195; 1996: 88-112; 
Vinogradova 2000: 93, 105, fig. 20). The settlements are located on the moun- 
tain sides in the system of the Kyzylsu river. Karim-Berdy measures 500 x 300 
m. Here were discovered an oval stone hearth and a rectangular structure with 
walls constructed of horizontal stone rows up to 0.65m high. 

The first group - 60% of the ceramics - is wheel-made, a vessel fragment is 
red-and-brown, on the outside it is coated with a white slip. One finds here pots 
with a narrow mouth, deep bowls, pots with a biconical body and sometimes 
with an applied-roller, vessels with vertical walls and a complex-profiled ridge. 
The second group is represented by hand-made ceramics: pots with an applied- 
roller, vessels with a spout and round-bottomed pots. A special third group 
consists of wheel-made vessels, small in number, painted in red against a light 
background. 

In Kangurt-Tut's upper layer there are vessels with a spout and a handle- 
support, painted fragments, round-bottomed pots and large vessels (khum) with a 
truncated conical base, maintaining the traditions of the Namazga VI stage of the 
BMAC culture. 

In Karym-Berdy, a bronze celt-adze was found (Fig. 53: 26). Analogous 
articles are known in eastern Kazakhstan, in the Palatzy hoard an example of this 
types was found together with a dagger with a mushroom-shaped top belonging 
to the Karasuk culture and a bracelet with conical spirals typical of the Begazy 
and later Andronovo complexes (Chernikov 1960: pi. x: 3, 4; Chlenova 1972: pi. 
61: 8, 9). Tadzhikistan monuments belong to the wide range of cultures repres- 
ented in northern Bactria: Kuchuk II (Askarov and Al'baum 1979; Sagdullaev 
1985; Shirinov and Shaydullaev 1988), in southern Bactria: Tillya-tepe 
(Sarianidi 1989), in Margiana: Yaz I (Masson 1959) and on the whole are 
synchronous with the cultures with the applied-roller ceramics in Fergana: Chust 
(Zadneprovsky 1962), though in the latter' s complex there is only hand-made 
pottery, wheel-made vessels being absent. 

The chronology of the monuments is debatable. Western scholars assume the 
formation of the culture of the barbarian occupation period in the third quarter of 
the 2nd millennium BC (Kohl 1984: pi. 2; Hiebert and Lamberg-Karlovsky 
1992). Most of the domestic researchers accept the date of the early 1st 
millennium BC. As grounds for this date — no earlier than late 2nd-early 1st 
millennium BC — serve: 1) applied -roller ceramics similar not to the Alekseevka 
type but to the later Dongal-Nur type, which is substantiated stratigraphically; 2) 
types of metal artifacts finding a wide range of analogies from the Ukraine to 



290 



CHAPTER TWENTY-ONE 



Siberia where they were identified on the sites of the Karasuk culture and are 
known as far as China during the Zhou period. A large number of ceramic forms 
and types of metal artifacts were directly adopted in the Early Iron Age. 





o 

N 

1 

g 


1 


B 

(TO 

H 

g. 


H 
o 

tfQ 

N 


H 

tr 

1 

O 


00 

Ss* 
o 

"B- 

S. 


YazII-KobadI 




+ 






+ 




Yaz I - Tillya 


+ 


+ 


+ 


+ 






Applied-roller ceramics 


+ 








+ 


+ 


Bishkent 






+ 




+ 


+ 


BMAC 


+ 


+ 


+ 


+ 


+ 


+ 


Hissar 






+ 


+ 






Harappa 












+ 



Table 13: Stratigraphy of the settlements of Northern Bactria with Andronovo ceramics 



Site 


Trumpet- 
shaped 
earring 


Wide 
bracelet 


Large 

bronze 

beads 


Plaque, 
mirror 
with a loop 


Stone 
beads 


Earring 

with 

hook 


Muminabad 


+ 


+ 


+ 








Chakka 


+ 












Sazagan 


+ 












Dzham 


+ 




+ 








Dashti-Kozi 


+ 


+ 




+ 


+ 


+ 


Kumsay 




+ 








+ 


Tuyun 












+ 


Tandyryul 


+ 




+ 




+ 




Nurek 










+ 




Bustan 6 


+ 


+ 


+ 


+ 


+ 


+ 



Table 14: Co-occurring ornaments of the Andronovo and BMAC types 



CHAPTER TWENTY -TWO 

TRANS-CASPIA AND TURKMENIA 

The north-east coast of the Caspian Sea, connecting Europe and Asia, played a 
significant role in the history of Trans-European migrations. Now it is a desert. 
In ancient times the ecology of the zone was determined by seasonal trans- 
gressions of the Caspian Sea: when the water level in the sea rose, the level of 
the water table rose as well making the region suitable for seasonal pasturing of 
herds. Although palaeo-ecologists differ as to the chronology of these trans- 
gressions, one may presume that by the late 3rd-early 2nd millennium BC there 
had occurred the Turali transgression. To this period belongs a large number of 
sites of the Poltavka culture of the Volga region which are also found on the 
Mangyshlak peninsula (Kuz'mina 1976a; Vasil'ev (ed.), 1995 Galkin 1978, 
1998). There are no data about their advancement further south; attempts to 
assign to this culture the small collection of hand-made ceramics with herring- 
bone ornament found at the early settlement of Kelleli, Margiana (Masimov 
1979: 121, fig. 6: 30-36), appears insufficiently substantiated due to this complex 
being small and indistinctive. 

In the 2nd millennium BC the Caspian regression was underway, reaching its 
maximum in the last quarter of the 2nd millennium BC, when a xerothermal 
climate led to desiccation and cooling. This forced tribal movements further to 
the north and east from the Timber-grave-Andronovo contact zone of the Volga- 
Ural steppe. This route is demonstrated by chains of sites near springs and 
ancient wells (Maps 4a; 13; 18; Fig. 44). One group of sites is concentrated on 
the north coast of the Aral Sea near Aral'sk (Formozov 1947, 1951; Vinogradov 
1959; Vinogradov et al. 1973; Glikman et al. 1968). Other sites stretch in a series 
from the Emba along the Caspian eastern coast in the Sam desert, on 
Mangyshlak alongside the Ust Urt, abundant in springs, over the Krasnovodsk 
peninsula and the water-rich central Uzboy to reach the border of the desert near 
the farming oases of southern Turkmenistan (Kuz'mina 1963). Here begins a 
third group of sites, located alongside the border of the oases and desert 
eastwards over the lowland valley of the Kopet-Dag (Figs. 49; 50: 1-39). Further 
on in the north-eastern steppes ceramics have been found all the way down- 
stream of the Murgab as far as the delta (Fig. 50: 40-53). Odd collections are 
known from Kelifskiy Uzboy. At these sites there were found tanged stone 
arrowheads, a bronze javelin (Bala-Ishem) and a spear (Asgabat; Fig. 53: 3,4). 

The ceramics discovered at the sites are coarse, hand-made and represented 
by jars, pots with a rounded shoulder and a broad neck pots with a ledge on the 
shoulder, sometimes vessels with a narrow cylindrical neck. The vessels lack 
decoration or are decorated only in the upper section or only along the rim or the 
shoulder. There ornament is executed in plain stamp or incized. The absolutely 
predominant pattern is vertical, more rarely horizontal, herring-bone. The disco- 



292 CHAPTER TWENTY-TWO 

ery of fragments with triangles hatched in bands and impressions underneath the 
rim are singular. Though absent at a number of sites, ceramics with an applied- 
roller constitutes an imposing group (Figs. 49: 29; 50: 8, 12, 14, 23, 24). 

The ceramic complex from the sites is not distinctive enough to assign it to 
the Andronovo or Timber-grave culture. It is more appropriate to use the cover 
term 'sites of the steppe type' (Kuz'mina 1964: 147; Mandel'shtam 1966a: 243). 
The regional and chronological distinctions are traced only in the ceramics of the 
north Caspian downstream of the Uil and Sagyz rivers, and the north Aral Sea 
where we often find vessels with a ledge on the shoulder and the richer Alakul' 
type decoration. It may be possible to distinguish a special variant of Murgab 
ceramics (Fig. 82). The complex of early Andronovo-Kozhumberdy type pottery 
was found in a grave at Yakhirbay 3 (Cattani 2004, fig. 5) and Togolok 1 . 

The criterion for establishing the chronology of the complexes is the presence 
of the applied-roller, characteristic of the pottery of the steppes of the Late 
Bronze Age (13th 1 1th centuries BC). This conclusion is supported also by the 
stratigraphy of the settlements of the Namazga VI period. On arrival at the 
foothill territory pastoralists came into contact with the neighboring farmers and 
their vessels appeared at the settlements. In accordance with the stratigraphy of 
the settlements two stages of interaction are identified: the late third quarter of 
the 2nd millennium BC and the 13th (12th)— 1 1 centuries BC, the upper date 
being the 1 1th century BC, when the Yaz I culture emerged in Turkmenia. 

Ceramics without the applied-roller were found only at the settlement of 
Anau on the southern hill in Layer C (Namazga VI), overlain by Layer D (Yaz I) 
(Fig. 52: 4,5,10,18,20,22; Pumpelly 1908: 142, 143, pi. xv,7-9). In the upper sec- 
tion of the Namazga VI layer at Tekkem-depe, pottery pieces were found with a 
ledge on the shoulder and incised ornament in the form of triangles between the 
rows of oblique notches (Fig. 71: 1-8, 14; Shchetenko 1999: 109; 1999c: fig. 2; 
Fig. 71: 1-8). These vessels may belong to either the Alakul' or the Alekseevka 
type. At the other sites of the foothill territory — El'ken-depe, Namazga High 
point (period 5-7), Sermancha, Tekkem (Kuftin 1954: 25; Marushchenko 1959: 
60-62, fig. v; Ganyalin 1956: 3; 1959; Shchetenko 1972: 530; 1999a: 108-110; 
1999b: 323-335) — vessels were found only in the overlying layer (Fig. 52). 

At Tekkem underneath the burnt layer which ended the settlement of the 
Namazga period, garbage layers were revealed with new pise houses abandoned 
again by the farmers. Later the ruins were occupied by pastoralists who erected a 
rectangular dwelling of the steppe type with vertical posts (Ganyalin 1959). In 
the house and the surrounding premises there were found animal bones as well as 
stone maces and pestles, including those with cones, whose analogies are well 
known, among other cultures at Namazga. The ceramics discovered in the over- 
lying layer of the farmers' settlements was represented by hand-made vessels, 
occasionally made to a textile pattern, of pot/jar shape. Ornament was on 38% of 
the pots: crosses, oblique notches, herring-bones, and pearls over the rim, and 
now and then herring-bone and triangles, including fringed ones, over the shoul- 
der. A number of the vessels are decorated over the shoulder with applied-rollers 
with notches. It is a classical complex of the Alekseevka type that determines the 
date of the pastoralists' occupation of the settlements of southern Turkmenia. 

Of the same date are the metal articles - double-edged knives with a guard, 
single-bladed knives (Namazga, Tekkem), a fragment of a mold for casting a pin 



TRANS-CASPIA AND TURKMENIA 293 

and double-edged knife from Namazga VI (Fig. 71: 13), as well as a metalwor- 
ker's assemblage from Tekkem: a furnace, three stone molds (Fig. 71: 12) and a 
blank mold lids, ore, an anvil, a stone axe, a bracelet, a fragment of a single -bla- 
ded knife and plate with a loop. The molds are designed for casting knives with a 
guard and plates with loops (Shchetenko 1999a: 109; 1999b: 323-5, fig. 2-4). 

All these articles are typical of the wide steppe zone of the Late Bronze Age. 
The findings at Tekkem and Namazga prove that the pastoralists, having arrived 
from the north, maintained in the south along with cattle-breeding the other 
pivotal branch of their economy, metalworking. 

On the Murgab and its delta the pastoralist's pottery (Figs. 52; 71) was found 
in great numbers at the settlements of Auchin, Takhirbay 1,3,13,15, Gonur I, 
Adzhi-Kuyu 8, Kelleli, Togolok 2,3,8,12,13.17,21,31, possibly Taip (Masson 
1959; 27, 116; pi. xi; Sarianidi 1975, 1978: 549; 1977: fig. 65, 66; 1990: 54, 55, 
1 10, plate xii: 14, 15; liv; Kuz'mina et al. 1979). A site was explored near Gonur 
(excavations by Hiebert 1989). 

What were the interrelations between the steppe and the farming population? 
V. Sarianidi presumes that the farmers peacefully co-existed with the cattle-bree- 
ders. In contrast, A. Ganyalin (1956), A. Marushchenko (1959), E. Kuz'mina 
(1963: 154), and A. Mandel'shtam (1966a: 242) hold that the cattle-breeders 
would intervene in the oases and war was one of the reasons for the end of some 
of the settlements. It is demonstrated by the appearance of fire on the settlement 
of Tekkem whose ashes were occupied by the peoples from the steppes. 

More likely, the character of interaction would vary and the exchange of 
cattle and metal for grain and craft articles would alternate with a period of 
hostilities. It is this character of interaction between the pastoralist and farming 
tribes that was typical throughout all the subsequent history of Central Asia. 

The downfall of the Namazga VI culture makes it possible to suggest that the 
internal crisis of the farming culture facilitated the conquest of settlements and in 
some cases the pastoralists would arrive at the already deserted towns. 

What is the cultural origin of the new-comers? The features characteristic of 
the early Andronovo (Novokumak, Alakul', Fedorovo, Kozhumberdy) and early 
Timber-grave (Berezhnovka, Pokrovskiy) ceramics are absent in Turkmenia. Few 
fragments are comparable with the late Alakul' and Timber-grave but the absolu- 
te majority of the pottery no doubt goes back to the Late Bronze Age and is rela- 
ted to the late Timber-grave pottery of the Ivanovo type (Agapov et al. 1983) and 
to the late Andronovo vessels of the Alekseevka type, which combines ceramics 
with an applied-roller on the shoulder with vessels ornamented poorly, and pre- 
dominantly with herring-bone, or completely lacking decoration. In the collec- 
tions that I am familiar with, the Nur and Dongal types of ceramics of the 11 th - 
9 th centuries BC, the applied-roller technique under the rim is not typical. It is no 
wonder since in this period in Turkmenia there already existed the Yaz I culture. 

It should be stressed that in the period of the applied-roller ceramics the 
steppe culture from the Danube to the Altai Mountains was uniform due to the 
increased mobility of the tribes and the establishment of active latitudinal ties. 
(The dissenting point of view (Kutimov 1993: 314-322) seems unfounded to me 
and is invalidated by the conclusions of V. Otroshchenko (2002)). 

Instrumental in establishing whence came the massive stream of the 
pastoralist population into southern Turkmenia is the analysis of the cemeteries. 



294 CHAPTER TWENTY-TWO 

At Ust'-Yurt in Sarakamanzhastik two kurgans were discovered. The mounds are 
formed of earth and stones; the burial was placed in a stone cist in the flexed po- 
sition on the left side, head to the east, with an unornamented vessel. At Mangy- 
shlak in Ayrakty one flat grave was explored, the dead lying flexed, head to the 
south, with fragments of a vessel ornamented in plain stamp (Galkin 1998: 20, 
21). In the Bol'shiye Balkhany desert archaeologists have explored the cemete- 
ries of Patma-say and Karalemata-say, in the foothills of the Kopet-Dag west of 
Kyzyl-Arvat the cemeteries of Parau I and II (Fig. 47: 1, 2, 9, 12, 13; Mandel'- 
shtam 1966a: 240-242; fig. 52; 1966b: 105-108; fig. 39, 40). These cemeteries 
consist of several mounds with a stone bank, 6-10m in diameter and 0.7-lm in 
height, in one case with a stone circle. The grave pit is formed of earth or has a 
stone cist overlaid with stone slabs on which are placed coals and, in one case, a 
bull's skull. The dead lie in the flexed position on the left side, head to the east or 
north-east; in Gyzylgy-kum the skeleton is burnt. Each grave contains one jar- 
shaped vessel, one of them (Patma-say; Fig. 47: 9) is decorated with impressions 
underneath the rim, another (Parau) — with zigzags; the rest are unornamented. 
The stone structures and the role of fire in the rite point to Andronovo influence. 
However, the eastern orientation gives grounds to relate the cemeteries to the late 
Timber-grave culture. The anthropological type of the newcomers is proto- 
Europoid, characteristic of the pastoralists of the Eurasian steppes. The same is 
true for the skull from Takhir-bay 3 (Ginzburg and Trofimova 1972: 69, 70). 

Thus the settling of Trans-Caspia and western Turkmenia in the 12th 1 1th 
centuries BC was realized mainly by the Timber-grave population who, as the 
migration progressed, incorporated part of the western Andronovans (corporative 
migration). The migration went from the west through the Timber-grave- 
Andronovo contact zone on the left bank of the Volga and southern Ural. One 
migration route went along the Emba; another — from the midstream of the Ural 
to the west of the Aral Sea, then further alongside the Uzboy and to the south. 
On the Murgab the immigration may have followed the central Amu-Darya. The 
cartography of the sites and the superposition of the maps of the later seasonal 
migrations demonstrate their coincidence. Hence, pastoralists as early as the 
12th 1 1th centuries BC developed the traditional routes of nomadic migrations 
that functioned up to the 19th century (Vinogradov et al. 1973: 102, 103; 
Gorbunova 1993-94: map on p. 2). 

The topography of the sites in the deserts of Trans-Caspia near the wells per- 
mits us to conclude that already in the early fourth quarter of the 2nd millennium 
BC there formed not only seasonal transhumance but also the fully nomadic type 
of cattle-breeding. Its prerequisites would slowly take shape in the culture of the 
Andronovo and Timber-grave tribes, who developed the skills of specialized 
cattle-breeding, invented yurts and wells. The farmers of Central Asia were not 
familiar with these innovations. Their ceramics are never found in the desert and 
there is absolutely no evidence of their migration to the north and participation in 
the ethnogeny of the Iranian-speaking Saka people and Scythians, whereas the 
path marked by the steppe sites reflects the actual route along which the 
migration of the cattle-breeding tribes to the south was realized, from the Volga 
and Urals, where one should seek its genesis, since the cultural elements 
represented in Turkmenia had there a centuries-old history of preceding 
development. 



PART THREE 

THE GENESIS OF THE DIFFERENT BRANCHES 

OF INDO-IRANIANS 



CHAPTER TWENTY-THREE 

MODERN STATE OF THE PROBLEM OF INDO-IRANIAN ORIGINS 

The Introduction provides the opposing views of Indo-Iranian origins that have 
been voiced in linguistics and archaeology over the course of more than two 
centuries. During the last decade since the completion of the earlier Russian 
edition of this book, discussion has continued unabated. 4 

A triumph of Russian Indo-Iranian studies was the international symposium 
of 1977 in Dushanbe on "Ethnic problems of the history of Central Asia in the 
early period". Among its participants were leading linguists, historians and 
archaeologists: I. D'yakonov, V. Abaev, V. Livshits, I. Steblin-Kamensky, G. 
Bongard-Levin, B. Litvinsky, E. Grantovsky, I. Aliev, M. Pogrebova, K. 
Smirnov, E. Kuz'mina, V. Sarianidi, V. Gening, A. Askarov, I. Masimov, the 
anthropologist V. Alekseev and others. The general thrust of these studies was 
the localization of the Indo-Iranian homeland in the steppes and their subsequent 
migration to Central Asia (Asimov 1981: 44-52). 

Also in attendance were Indian scholars, S. S. Misra, B. B. Lai, B. K. Thapar, 
R. C. Gaur, L. Gopal, and A. H. Dani (Pakistan), and European researchers, B. 
and R. Allchin, R. Ghirshman, K. Jettmar and V. Brentjes. The idea of an Indo- 
Iranian migration from the north predominated and the Aryans' culture after their 
arrival in India was correlated with the Painted Gray Ware culture. 

The establishment of this hypothesis for an Indo-Iranian migration was a 
break-through in Russian science which had for years labored with the concept 
of autochthonous development. The symposium of 1977 brought euphoria. The 
Indo-Iranian attribution of the Timber-grave and Andronovo cultures received 
universal recognition. 

However, 1980 saw the beginning of a heated discussion about the new 
concept formulated already in 1972 by the prominent linguists T. Gamkrelidze 
and V. V. Ivanov (1980; 1984). Assuming connections between the IE languages 
and those of the Caucasus and ancient Near East, they moved the original 
homeland to 4th millennium BC Eastern Anatolia, whence the Indo-Aryans (IA) 
went to Mitanni and India, and the Iranian Scythians, not until in the 8th century 
BC passed through Central Asia northwards into the steppes following the other 
Indo-Europeans. This hypothesis was dismissed by Soviet archaeologists. Many 
Near -Eastern borrowings were called into question by I. D'yakonov (1980). 

Proceeding from completely different considerations, C. Renfrew in 1987 
localized the Proto-Indo-Europeans in Anatolia, a center of inception of the 



The chapter was finished in 2004, and I could not here make full use of a number of 
important recent publications: Bongard-Levin 2003; Scripta gregoriana (Sbornik v chest' 
G. M. Bongard-Levina), Moscow, 2003; Tsentral'naya Aziya. Istochniki, istoriya, 
kul'tura. Moscow, 2003; U istokov tsivilizatsii, Moscow, 2004; Transoxiana 2004; 
Fussman et al. 2005; Kogan 2005. 



298 CHAPTER TWENTY-THREE 

Neolithic economy from where they passed in the 7th-6th millennium BC 
through the Balkans to settle in Europe bringing along farming and cattle- 
breeding skills. In doing so, according to his Model A, they immediately went 
eastwards to India, while according to Model B the original homeland of all the 
Indo-Iranians was localized in the steppes whence they later moved into Iran and 
India. 

C. Renfrew's critics pointed out that the distribution of cultural innovations is 
often conditioned not by migration of a new population but by cultural 
borrowings. 

A. and S. Sherratt (1988) expressed an alternative opinion holding that IE 
settlement and the Anatolian-Pontic interaction took place not in the 6th 
millennium BC but only after the secondary products revolution of the 4th 
millennium BC. In 1990 I, M. D'yakonov (D'yakonov 1990: 53-65) also placed 
the original homeland of the pre-Proto-Indo-Europeans in the 6th millennium BC 
in the Near East believing that S. Starostin had established ancient ties with the 
Caucasian languages and those of the Near East. He assumed a migration of the 
Proto-Indo-Europeans through the Balkans and Danube and linked it with the 
distribution of Linearbandkeramik culture. The Indo-Europeans continued to 
develop in Europe, and as for the Indo-Iranians, I. M. D'yakonov (1995: 123- 
130) acknowledged them to be the creators of the Andronovo culture linking 
their migration with the spread of this culture over the south of Central Asia. 

In 1989 J. P. Mallory published In Search of Indo-Europeans, in which he 
most strictly and with much reasoning advocated the concept, expressed as early 
as the 19th century, of the localization of the IE original homeland in Europe, 
underlining the role of the Pontic steppes, the place of domestication of the 
horse. 

In 1990 in Delhi, a conference was held at which a group of nationalistically- 
charged intelligentsia declared that the hypothesis of the Aryans' migration was 
created by imperialists, whereas India was the original homeland of the Indo- 
Aryans and that these were the founders of the high civilization of Harappa. This 
hypothesis is widely discussed and very popular in today's India. 

What is the state of the problem of Indo-Iranian origins today? Laying 
absolutely no claim to the comprehensiveness of analysis of this question I will 
focus only on a few points that are significant from my perspective. 



Over recent years the proceedings of several international conferences have 
been published as well as some fundamentally important monographs: T. Ya. 
Elizarenkova (1995), T. Ya. Elizarenkova and V. N. Toporov (1995), E. A. 
Grantovsky (1998, the book was prepared by S. V. Kullanda), O. N. Trubachev 
(1999). 

There are still three principal hypotheses under discussion. Whereas the 1996 
Philadelphia conference (Mair 1998) was marked by the debate between J. P. 
Mallory (1998) and C. Renfrew, at the conference of 1999 at Arkaim C. Renfrew 
made a remark that t